# Nowhere to hide: pumas, black bears, and competition refuges

> In Colorado and California, black bears found and took over many puma kills in summer, and pumas could find nowhere on the landscape to hide them. By reading tracks and signs at GPS-located kill sites and watching kills on video, researchers found that pumas made up for the stolen meat by killing more often, with kill rates up by nearly half in the bear season.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2015-elbroch-lendrum-nowhere-hide-pumas-black/

## Citation and access

- **Authors:** L. Mark Elbroch, Patrick E. Lendrum, Maximilian L. Allen and Heiko U. Wittmer
- **Year published:** 2015
- **Journal or source:** Behavioral Ecology
- **Full citation:** Elbroch, L. M., Lendrum, P. E., Allen, M. L. and Wittmer, H. U. 2015. Nowhere to hide: pumas, black bears, and competition refuges. Behavioral Ecology 26(1): 247–254.
- **DOI:** 10.1093/beheco/aru189
- **Access status:** Publisher link only
- **Where to find it:** https://doi.org/10.1093/beheco/aru189

## Tags

- **Species:** Mammals > American black bear; Mammals > Mountain lion
- **Region:** North America > United States
- **Tracking:** Bones; Feeding signs; Hair; Kill sites; Scat; Tracking; Tracks; Tracks and signs identification
- **Behaviour and ecology:** Behaviour; Carnivore coexistence; Competition; Feeding; Hunting; Interspecific competition; Intraguild interactions; Kleptoparasitism; Predation; Scavenging
- **Methods and technology:** Akaike Information Criterion (AIC); Camera traps; CyberTracker; CyberTracker Tracker Certification; Global Positioning System (GPS) collars; GPS cluster analysis; Logistic regression; Resource selection function
- **Data type:** New field data
- **Observer qualification:** Certified trackers
- **Discipline:** Behaviour and ecology; Wildlife tracking
- **Year published:** 2015

## Detailed keywords

- **Species:** Studied: puma (Puma concolor) and American black bear (Ursus americanus). Prey: in Colorado, elk (Cervus elaphus), including calves, and mule deer (Odocoileus hemionus), including fawns, plus smaller prey such as American beavers (Castor canadensis) and North American porcupines (Erethizon dorsatum); in California, black-tailed deer (O. h. columbianus), including fawns, the only ungulate prey, plus black-tailed jackrabbits (Lepus californicus), northern raccoons (Procyon lotor) and California ground squirrels (Otospermophilus beecheyi). Other scavengers regularly detected at puma kills: coyotes (Canis latrans), golden eagles (Aquila chrysaetos) and common grey foxes (Urocyon cinereoargenteus) in Colorado; common grey foxes, coyotes, turkey vultures (Cathartes aura) and common ravens (Corvus corvax) in California. Examples from other studies discussed: brown bears and Eurasian lynx, Andean condors, spotted hyenas, lions, African wild dogs, cheetahs, leopards, tigers, wolves and moose.
- **Location and habitat:** Two study areas in the United States. (1) Western Colorado near the town of De Beque, about 1100 km², about half managed by the High Lonesome Ranch, with interspersed Bureau of Land Management land: rugged valleys at 1500 to 1700 m, canyons and plateaus up to 3000 m; pinyon-juniper woodland, Gambel oak, rangeland shrubs, mixed conifer and aspen forest; 295 mm mean annual precipitation, mostly snow. (2) Northern California, about 1000 km², almost entirely within the Mendocino National Forest, 400 to 2450 m: oak woodland, dense chaparral, grassland, mixed conifer and fir forest, shrubs and meadows; 1320 mm mean annual precipitation.
- **Topics:** kleptoparasitism (food stealing) by a dominant competitor; interference competition; competition refuges (spatial and temporal); carnivore coexistence; kill rates; handling time and search time; partial prey consumption; optimal foraging versus kleptoparasitism; satiation; seasonal (“bear” and “no-bear” seasons) foraging; prey size; carrion subsidies to scavengers; indirect effects of competitors on prey; community framework for predation studies; puma killing a black bear that had displaced her.
- **Statistics or models:** Generalized linear mixed models (binomial with logit link, and Poisson with log link, individual puma as a random effect; SAS 9.3); model selection with Akaike’s Information Criterion corrected for small samples (AICc) and Akaike weights; resource selection functions using conditional logistic regression stratified by individual puma; correlation matrix to check collinearity (|r| > 0.5); covariates elevation, slope, aspect, vector ruggedness measure (VRM), distance to nearest river and habitat type. Possible inconsistencies in the paper: (1) in Colorado the text reports 437 predation events, but the prey listed (337 ungulates plus 106 smaller vertebrates) add up to 443; (2) in California 352 prey are reported, but 288 deer plus 62 small prey make 350; (3) the season effect on handling time is given the same statistic as the season effect on prey size (F1,750 = 42.68), which may be a copying error; (4) Table 2 (the model comparison) is referred to in the text but does not appear in the article PDF.
- **Data type:** New field data: 11 pumas captured in Colorado (February 2010 to March 2012) and 7 in California (December 2009 to November 2012), GPS clusters investigated in the field, and video cameras at active kills.

## Synopsis

### 1. What they did

Pumas are top predators, but where they live alongside American black bears, bears are the dominant competitors and often take over puma kills. This is called kleptoparasitism (food stealing). One idea for how competing carnivores coexist is that the weaker species uses “competition refuges”: places or times where they are less likely to meet the stronger one. The researchers asked how black bears affect the way pumas hunt and eat.

They worked in two study areas in the western United States: western Colorado and the Mendocino National Forest in northern California. They captured 18 pumas (11 in Colorado, 7 in California) and fitted them with Global Positioning System (GPS) collars that took a location every 2 hours. Field teams visited clusters of locations to find and identify kills, and placed video cameras at kills where pumas were still feeding. Bear presence at kills was recorded from video footage, footprints, scats and other signs.

Each study area was split into a “bear season” (when bears were active and found at puma kills) and a “no-bear season”. The researchers then tested three questions:

- Season and site. Did prey size, handling time (time spent at a kill), search time (time between kills) and kill rates differ between seasons and study areas?

- Why pumas leave meat behind. Was the amount of time pumas spent at large kills (40 kg or more) in the bear season better explained by optimal foraging (leaving when prey are easy to catch) or by bears taking the kill?

- Hiding places. Did pumas make kills in places where bears were slower to find them? They compared kills found by a bear within 48 hours with kills found after 96 hours or more, or never.

### 2. What they found

Kills and bears

- In Colorado, the teams investigated 1001 GPS clusters and documented 437 kills and 8 scavenging events. In California they investigated 598 of 609 clusters and identified 352 prey and 4 scavenging events.

- In the bear season, black bears were detected at 47.8% of puma kills of 40 kg or more in Colorado, and 77.2% in California.

- Cameras showed bears displacing pumas at 29 kills in each area. Pumas came back only once to 3 of those kills in Colorado (10%) and 8 in California (28%).

- Once in California, an adult female puma returned and killed a 3-year-old female bear that had pushed her off her kill.

Seasons

- In the bear season, pumas killed smaller prey, spent less time at each kill (43.2 hours on average, against 89.1 hours in the no-bear season) and searched for less time between kills (69.4 hours against 98.7 hours).

- Kill rates were higher in the bear season: 1.48 against 0.84 animals per week, and 1.14 against 0.77 ungulates (hoofed animals) per week. This is 77% and 48% more. These rates did not differ between the two study areas.

- Measured in kilograms per day, kill rates were the same in both seasons, but higher in Colorado (9.0 kg per day) than in California (6.0 kg per day).

Why pumas leave meat behind

- The best model explained time at a kill by prey size, the search time before the kill, and whether a bear was present. Pumas stayed longer at bigger prey and after longer searches, and left sooner when a bear came.

- After kills where a bear was present, pumas made their next kill sooner.

- This pattern does not fit optimal foraging. It fits bears driving pumas off their kills.

No hiding places

- None of the landscape features tested (elevation, slope, aspect, ruggedness, distance to the nearest river or habitat type) predicted how quickly a bear found a puma kill. Kills found quickly and slowly were often close together. The chance of a bear finding a kill was the same across the landscape.

### 3. What they concluded

- Black bears strongly affect how pumas feed. Pumas did not use spatial competition refuges to avoid bears. Instead, they appear to make up for lost meat by killing more often.

- The only relief pumas had from bears in both areas was a time refuge: winter, when bears hibernate.

- Higher summer kill rates are not caused by bears alone; prey such as newborn ungulates are more available then, and carcasses spoil faster in the heat. Pumas may also choose smaller prey in summer, which they can eat quickly before bears arrive.

- Because other scavengers also feed at puma kills, the effects of food stealing on pumas are likely underestimated.

- A dominant competitor may indirectly affect prey populations by making a top predator kill more. Predation can only be understood within a community framework, and future studies should separate the effects of predators and competitors on prey.

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

No trailing was reported. Tracks and signs identification was central to the field work at GPS clusters:

- Prey identification: teams searched each cluster for prey remains, including hair, skin, rumen and bone fragments.

- Kill or scavenge: the state of the remains, the presence and position of bite marks (throat, nose or back of the head indicating a kill) and which body parts had been eaten were used to decide whether the puma had killed the animal or was scavenging.

- Bear presence: bears at puma kills were detected from video footage, footprints, scats and other signs. These detections underpin all the main results.

### 5. Tracker or observer certification

93% of all field investigations of clusters were carried out by CyberTracker certified observers (citing Elbroch et al. 2011), which the authors say ensured consistent effort and expertise across both study areas. The level of certification is not given. Several people are thanked for their help with kill site investigations.

### 6. Other technologies, skills or innovations

- GPS collars (Lotek 7000SAW with Argos uplink, or Lotek Iridium 2D) taking locations every 2 hours.

- GPS cluster analysis to find kill sites, with locations loaded onto handheld GPS units to find clusters in the field.

- Motion-triggered video cameras (Bushnell Trophy Cam) at active kills, which recorded bears displacing pumas.

- Bear and no-bear seasons defined from the field data (the earliest and latest dates bears were found at kills), rather than calendar seasons.

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

- Kleptoparasites may indirectly affect prey populations by raising the kill rates of top predators. This may matter especially for rare prey in systems with many prey species.

- Puma kills stolen by bears may in turn feed bear populations, and bears also prey on deer fawns. Whether this increases or buffers predation on ungulates is still an open question.

- Predation studies should look at the whole community: predators, prey, carrion, scavengers and competitors together.

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

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

- The study rests on reading kill sites: telling a kill from a scavenge by bite marks and remains, and detecting bears from footprints and scats where cameras were not placed. Having most of this work done by certified observers gave the researchers a consistent standard across two study areas.

- GPS shows where a puma stopped; tracks and signs at the site reveal what happened there and who else came to feed.

## Cite this summary

Lawrence, K. 2026. Summary of: Nowhere to hide: pumas, black bears, and competition refuges (Elbroch and Lendrum et al., 2015). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2015-elbroch-lendrum-nowhere-hide-pumas-black/