# Bed site selection by a subordinate predator: an example with the cougar (Puma concolor) in the Greater Yellowstone Ecosystem

> Where does a cougar sleep when wolves and bears share its home? By following GPS collars to 599 cougar beds in the Greater Yellowstone Ecosystem, certified trackers found that cougars pick sunny, south-facing slopes in winter but favour steep, hidden spots near forest edges, the kind of places that make escape easier. Even a top predator, it seems, beds down like prey.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2017-kusler-elbroch-bed-site-selection-subordinate/

## Citation and access

- **Authors:** Anna Kusler, L. Mark Elbroch, Howard Quigley and Melissa Grigione
- **Year published:** 2017
- **Journal or source:** PeerJ
- **Full citation:** Kusler, A., Elbroch, L. M., Quigley, H. and Grigione, M. 2017. Bed site selection by a subordinate predator: an example with the cougar (Puma concolor) in the Greater Yellowstone Ecosystem. PeerJ 5: e4010.
- **DOI:** 10.7717/peerj.4010
- **Access status:** Open access (CC BY 4.0)
- **PDF:** https://www.originalwisdom.com/wp-content/uploads/2026/11/2017_Kusler_and_Elbroch_etal_bed-site-selection-by-a-subordinate-predator-an-example-with-the-cougar.pdf

## Tags

- **Species:** Mammals > Mountain lion
- **Region:** North America > United States
- **Tracking:** Beds; Hair; Kill sites; Tracking; Tracks and signs identification
- **Behaviour and ecology:** Behaviour; Habitat selection; Interspecific competition; Intraguild interactions; Predation risk; Predator avoidance
- **Methods and technology:** Akaike Information Criterion (AIC); Chi-square test; Cover boards; CyberTracker; CyberTracker Tracker Certification; Geographic Information System (GIS); Global Positioning System (GPS) collars; GPS cluster analysis; Logistic regression; Resource selection function; Terrain ruggedness index
- **Data type:** New field data
- **Observer qualification:** Certified trackers
- **Discipline:** Behaviour and ecology; Wildlife tracking
- **Year published:** 2017

## Detailed keywords

- **Species:** Studied: cougar (Puma concolor), nine collared cougars (three males, six females). Dominant competitors in the study area: wolf (Canis lupus), American black bear (Ursus americanus) and grizzly bear (Ursus arctos horribilis). Other carnivores present: coyote (Canis latrans) and red fox (Vulpes vulpes). Ungulate prey present: elk (Cervus elaphus), mule deer (Odocoileus hemionus), white-tailed deer (O. virginianus), Shiras moose (Alces alces shirasi), bighorn sheep (Ovis canadensis) and North American pronghorn (Antilocapra americana). Trailing hounds were used to make cougars retreat to a place where they could be safely approached for capture.
- **Location and habitat:** About 2,300 km² of the southern Greater Yellowstone Ecosystem, Teton County, Wyoming, United States: parts of Bridger-Teton National Forest, Grand Teton National Park and the National Elk Refuge, northeast of Jackson. Elevations from 1,800 m in the valleys to over 3,600 m. Short summers and long, snowy winters (average summer 6.9 °C, average winter −7.2 °C, extremes to −36.7 °C). Foothill grasslands, big sagebrush shrub-steppe, Douglas-fir and aspen forests, higher coniferous forests (lodgepole pine, subalpine fir, Engelmann spruce, whitebark pine), and cottonwood and willow along rivers. Beds were often in cliff bands, talus fields and boulder fields.
- **Topics:** bed site selection; resting and sleeping sites; refugia and “competition refuges” for subordinate predators; predator avoidance hypothesis; thermoregulation hypothesis; intraguild competition with wolves and bears; escape terrain; vegetative concealment; canopy cover; aspect and sun exposure; seasonal differences (winter December to May, summer June to November, based on elk migration); second-order (landscape) and fourth-order (microsite) selection; kittens hidden from predators.
- **Statistics or models:** Resource selection functions: logistic regression (binomial, logit link) comparing 599 beds with five random points each across the study area; generalized linear mixed models with cougar identity as a random intercept for the microsite data (140 beds and 50 random points per cougar per seasonal home range); model comparison with Akaike’s Information Criterion corrected for small samples (AICc), ΔAICc and Akaike weights; post-hoc combination models; correlation matrix and chi-square tests for collinearity; 95% fixed-kernel home ranges; vector ruggedness measure (VRM); ArcGIS 10.1 and Geospatial Modeling Environment. Inconsistencies in the paper: (1) the abstract says winter microsite beds supported both hypotheses, but the results say winter microsite beds supported thermoregulation over predator avoidance (Table 2 shows the thermoregulation model with the lowest AICc, 214.8 versus 230.3 for the combination model); (2) the results report steeper slope as significant in the summer microsite combination model, but the combination model shown in Table 2 (canopy + conc + topo + on_feat) has no slope term; (3) the results name benches and cliff bands as the selected summer topography, while the discussion says sloping hillsides and cliff bands; (4) the methods list “seven” microsite habitat types but name six; (5) escape terrain is described as a measured distance in the methods but as a within-200-m yes or no variable in Table 2; (6) a results subheading says “den sites” where bed sites are meant, and the Table 2 caption says “landscape level” for the microsite table.
- **Data type:** New field data: GPS clusters visited from 23 December 2012 to 19 January 2016 (1,718 clusters: 599 beds, 754 kills and 366 sites with nothing found). Microsite data at 140 beds (80 winter beds from eight cougars, 60 summer beds from six cougars). Raw data published as supplemental files.

## Synopsis

### 1. What they did

Where an animal sleeps matters, because sleep is one of its most vulnerable times. Studies of prey animals such as deer and primates show that bed sites help them stay warm or cool and stay safe from predators. Far less is known about predators that are themselves at risk from bigger competitors. In the Greater Yellowstone Ecosystem, cougars are subordinate to wolves, black bears and grizzly bears: bears steal their kills and force them off their prey, while wolves affect which prey cougars take and where they live, and kill cougars, including kittens.

The researchers asked whether cougars choose bed sites that help with temperature (the thermoregulation hypothesis), with avoiding dominant predators (the predator avoidance hypothesis), or both.

- Finding beds. Cougars were caught each winter with the help of hounds and fitted with satellite Global Positioning System (GPS) collars taking 12 to 24 locations a day. Clusters of locations were picked out in Google Earth and visited in the field.

- Identifying beds. A bed was a circular depression in vegetation or snow with identifiable cougar hair. To keep beds separate from feeding, only beds more than 500 m from known kills, and outside times spent handling prey, were used.

- Landscape scale. 599 beds were compared with random points across the study area for habitat type, distance to forest edge, slope, ruggedness, elevation and aspect (which way the slope faces).

- Microsite scale. At 140 beds and at random points within each cougar’s seasonal home range, the team measured canopy cover, how well the spot was hidden (with a concealment board viewed from 10 m), nearby physical features such as trees, cliffs or boulders, habitat, topography, slope and distance to “escape terrain” (cliff bands, talus fields and downed wood).

- Testing the ideas. For each season and scale, they built sets of models for each hypothesis and compared them, along with a combined model.

### 2. What they found

The field effort

- From December 2012 to January 2016 the team visited 1,718 clusters and documented 599 beds, 754 kills and 366 sites with nothing found.

- The 599 beds came from nine cougars (three males, six females): 312 in winter and 287 in summer.

Landscape scale

- Winter: beds supported both hypotheses. Cougars bedded at lower elevations, on steeper slopes and close to forest edges (on average about 57 m away). They avoided sagebrush, meadow and riparian habitats. They favoured south-facing slopes most, and also east and west-facing slopes.

- Summer: beds strongly supported predator avoidance over thermoregulation. Cougars bedded close to forest edges (about 76 m on average) and on steeper slopes, and avoided sagebrush and meadows. Aspect did not matter in summer.

Microsite scale

- Winter: the results favoured thermoregulation. Cougars chose spots with high canopy cover (about 87% on average) and in “rugged” habitat of cliff bands and talus fields, which the authors note likely also helped them avoid predators.

- Summer: beds supported both hypotheses. Cougars chose spots with high concealment (about 82%), high canopy cover (about 92%), on steeper slopes, and on benches and cliff bands.

- Most beds were within 200 m of escape terrain (77% in summer, 90% in winter), but cougars used escape terrain in proportion to how common it was, so the expected selection for it was not found.

### 3. What they concluded

- Just like prey species, a subordinate predator chooses bed sites that serve both temperature and safety. What mattered changed with season and scale.

- In winter, sun-facing slopes and lower elevations likely help cougars cope with the cold, while steep slopes near forest edges and avoidance of open ground, where wolves have the advantage, support escape. Winter is when cougars there have more negative encounters with wolves.

- In summer, safety seemed to matter more: bears roam widely and steal food and sometimes kill kittens, and most kittens are born in summer and are protected mainly by hiding them.

- Rather than choosing beds close to escape terrain, cougars may simply live in rugged, complex habitat when sharing ground with wolves.

- The sample was small (nine cougars at the landscape scale, eight at the microsite scale), and other populations and other subordinate predators need study.

- Bed sites may offer a new way to measure the “refuges” that let subordinate predators live alongside dominant ones. Researchers already visiting GPS clusters to study kills could record beds with little extra effort.

### 4. How tracking was or was not used

Tracks and signs identification was central to the study. Field teams visited GPS clusters and identified what had happened there: kills, beds, or nothing found. Beds were identified from signs: a circular depression in the vegetation or snow with identifiable cougar hair. Kill sites were also identified, so that beds linked to feeding could be left out. No trailing of cougar tracks by people is reported; hounds were used to trail cougars for capture.

### 5. Tracker or observer certification

98% of all cluster investigations were carried out by CyberTracker-certified observers (the paper cites Evans et al. 2009 and Elbroch et al. 2011), which the authors say ensured expertise and consistent field effort. The certification levels held are not given.

### 6. Other technologies, skills or innovations

- Satellite GPS collars (Lotek and Vectronic) taking 12 to 24 locations a day, with clusters picked out daily in Google Earth.

- Geographic Information System (GIS) layers: a land cover map reclassified into five habitat classes, a digital elevation model for slope and aspect, a vector ruggedness measure, and a mapped forest edge.

- Field microsite tools: a spherical crown densiometer for canopy cover, a 1 m by 50 cm concealment board viewed from 10 m in each direction, and a rangefinder for distance to escape terrain.

### 7. Implications for conservation (from the paper)

- Refuges are a key resource for subordinate predators as well as for prey, and deserve more research attention alongside studies of how prey shape predator movements.

- Bed sites that reduce the risk from competitors may save cougars energy and increase their fitness, since fleeing competitors is energetically costly.

- Bed site data could help researchers understand the habitat needs and energy use of subordinate carnivores in systems with several competing predators.

### 8. Implications for tracking (Original Wisdom's observations)

_These are Original Wisdom's observations, not the authors'._

- This study rests on reading signs at GPS clusters: telling a bed from a kill or from nothing at all, using the depression and the hair left behind. The authors chose certified observers for this work to ensure expertise and consistent effort.

- Of the 1,718 clusters visited, 366 (about one in five) gave nothing, a reminder that a GPS cluster shows where an animal paused but not what it did there.

## Cite this summary

Lawrence, K. 2026. Summary of: Bed site selection by a subordinate predator: an example with the cougar (Puma concolor) in the Greater Yellowstone Ecosystem (Kusler and Elbroch et al., 2017). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2017-kusler-elbroch-bed-site-selection-subordinate/