# Nuisance Ecology: Do Scavenging Condors Exact Foraging Costs on Pumas in Patagonia?

> Can a bird make a puma nervous? In Chilean Patagonia, certified trackers checked over 400 puma kills and found that pumas left half of their large kills after a single night, especially in open grassland where Andean condors gather. Pumas spent less time at kills where condors came, killed about 50% more meat per day than pumas in North America, and still hunted most in the risky open country where prey was plentiful.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2013-elbroch-wittmer-nuisance-ecology-do-scavenging/

## Citation and access

- **Authors:** L. Mark Elbroch and Heiko U. Wittmer
- **Year published:** 2013
- **Journal or source:** PLoS ONE
- **Full citation:** Elbroch, L. M. and Wittmer, H. U. 2013. Nuisance ecology: do scavenging condors exact foraging costs on pumas in Patagonia? PLoS ONE 8(1): e53595.
- **DOI:** 10.1371/journal.pone.0053595
- **Access status:** Open access (CC BY 4.0)
- **PDF:** https://www.originalwisdom.com/wp-content/uploads/2026/11/2013_Elbroch_and_Wittmer_nuisance-ecology-do-scavenging-condors-exact-foraging-costs-on-pumas.pdf

## Tags

- **Species:** Birds > Andean condor; Mammals > Mountain lion
- **Region:** South America > Chile
- **Tracking:** Bones; Caches; Feeding signs; Hair; Kill sites; Scat; Snow tracks; Tracking; Tracks; Tracks and signs identification; Trailing
- **Behaviour and ecology:** Behaviour; Feeding; Food caching; Habitat selection; Hunting; Interspecific competition; Intraguild interactions; Kleptoparasitism; Movement; Predation; Predation risk; Scavenging
- **Methods and technology:** Analysis of variance (ANOVA); Camera traps; Chi-square test; CyberTracker; CyberTracker Tracker Certification; Direct observation; Distance sampling; Geographic Information System (GIS); Global Positioning System (GPS) collars; GPS cluster analysis; Very high frequency (VHF) radio collars
- **Data type:** New field data
- **Observer qualification:** Certified trackers
- **Discipline:** Behaviour and ecology; Wildlife tracking
- **Year published:** 2013

## Detailed keywords

- **Species:** Studied: puma (Puma concolor), also called mountain lion, and Andean condor (Vultur gryphus), a large scavenging bird. Puma prey found at kills: guanaco (Lama guanicoe), huemul deer (Hippocamelus bisulcus, endangered), domestic sheep (Ovis aries), European hare (Lepus europaeus), Patagonian hairy armadillo (Chaetophractus villosus), culpeo fox (Lycalopex culpaeus) and upland goose (Chloephaga picta). Common in the study area but not studied: culpeo fox, southern caracara (Polyborus plancus), chimango caracara (Milvago chimango) and black-chested buzzard eagle (Geranoaetus melanoleucus). Comparisons discussed: pumas in North America with bears (Ursus spp.) and gray wolves (Canis lupus); brown bears (U. arctos) and Eurasian lynx (Lynx lynx); white-backed griffon vulture (Gyps africanus) for condor intake per feeding bout; mule deer (Odocoileus hemionus) and white-tailed deer (O. virginianus) for huemul growth rates.
- **Location and habitat:** Southern part of Chile’s Aysén District, just north of Lago Cochrane, central Chilean Patagonia, Chile, including Valle Chacabuco. Study area about 1063 km² of five habitats: valley steppe (74 km², flat and open, 200 to 600 m above sea level), mountain steppe (754 km², steep slopes, ravines and rock outcrops, 600 to 1,200 m), shrubs (138 km², dense woody understorey, 200 to 700 m), lenga (Nothofagus pumilio) forest (44 km², 700 to 1,200 m) and barren mountaintops (53 km², above 1,200 m). Prey biomass was highest in the open valley and mountain steppes and lowest in forests and on barren mountaintops. Note: the coordinates are printed as “W 47.8000°, S 72.0000°”, with the west and south labels apparently swapped; the figure caption gives the study area as 1,062 km², the text as about 1063 km².
- **Topics:** kleptoparasitism (food stealing) by scavengers; interference competition; predation risk and the landscape of fear; giving-up densities (GUDs) and a new measure, giving-up time (GUT), the hours a puma stays at a carcass it leaves with more than 10% of the edible meat; handling time and search time; kill rates; carcass abandonment; food caching; high-risk, high-reward foraging; optimal foraging theory; scavengers’ indirect effects on prey and communities; puma predation on endangered huemul; human persecution of pumas as an alternative explanation.
- **Statistics or models:** Generalized linear models (GLIM) with a Poisson distribution and puma as a random effect (handling time against prey weight, condor presence and caching; search time against handling time and prey weight); mixed-model analysis of variance (ANOVA) with puma as a random effect (search time, distances, kill rates, GUTs against risk class and prey density); ANOVA with Tukey-Kramer test (GUTs across risk classes); chi-square goodness of fit (kill locations against availability of risk classes); pairwise correlation (cut-off 0.5; canopy cover dropped because it correlated with condor presence); fixed kernel density (50%, 75% and 95% condor activity polygons, least squares cross validation) in ArcView 3.2 with the Animal Movement Extension; ArcGIS 9.1 and 9.3; SAS 9.2; square-root and log transformations. Prey densities from earlier direct counts and Distance sampling. Possible inconsistencies in the paper: (1) Table 3 gives the observed frequency of the 7 kills in risk category 2 as 0.3%, but 7 of 212 kills is 3.3% (and the column only adds to 100% with 3.3); (2) Tables 2 and 3 both use 212 kills but give different numbers of kills per risk class (for example 4 versus 7 in class 2, 60 versus 58 in class 7); (3) for the ungulate-only search time test the text reports F(1,245) = 2.13, P = 0.04, but an F value of 2.13 on those degrees of freedom would give a P value of about 0.15.
- **Data type:** New field data. Pumas captured March to September 2008 and 2009; nine pumas monitored for a mean of about 9.3 months. Kill sites were usually investigated within about 11 days (range 0 to 78) of the puma leaving; for one puma whose collar failed to transmit, sites were investigated 650 to 945 days later from data stored in the collar. Prey densities and habitat data come from the authors’ earlier work.

## Synopsis

### 1. What they did

Prey animals often give up food early when they feel at risk of being caught, and researchers measure this as “landscapes of fear”. The authors asked whether a top predator can face something similar from scavengers. In Chilean Patagonia, Andean condors gather in numbers at puma kills. The authors asked three questions: (1) Do condors cost pumas food? (2) If so, do pumas behave as if they know it? (3) Do pumas forage like prey animals in a risky landscape?

- Following pumas. They caught pumas in 2008 and 2009, and fitted nine with Global Positioning System (GPS) collars that took a location every 2 hours and sent the data by satellite (Argos). They monitored them for an average of about 9 months.

- Finding kills. Places where a puma made two or more locations within 150 m (GPS clusters) were visited by CyberTracker-certified observers, who identified prey, decided whether the puma had killed or scavenged it, noted whether it was cached, measured canopy cover, and recorded whether condors had been there.

- Measuring feeding. From the collar data they worked out handling time (hours a puma stayed with a kill), search time (hours until the next kill), distances moved between kills, and kill rates.

- Mapping risk. They gave each habitat a score for how easily condors could see a carcass, from forest (lowest) to open valley steppe (highest), and mapped where condors had found kills. Combining the two gave risk classes from 2 to 8.

- Giving-up time. For kills where a puma left more than 10% of the edible meat, they used the time spent at the carcass as a “giving-up time” (GUT): the shorter the time, the greater the risk the puma seemed to feel. They tested whether condor risk or prey density better explained GUTs.

### 2. What they found

Kills and condors

- They investigated 694 GPS clusters and found 433 kill sites and 6 cases of puma scavenging. Prey were 350 ungulates (302 guanacos, 41 domestic sheep, 7 huemul) and 83 small to medium-sized animals.

- Condors were recorded at 43% of 296 ungulate kills where their presence could be determined with certainty.

- Pumas left an estimated 29 kg of meat (range 0 to 75 kg) at each ungulate kill.

- Of 266 ungulate kills with known handling time, pumas left 133 (50%) after a single night with more than 10% of the edible meat remaining, and another 56 (21%) after two nights. The authors report that pumas never returned to check those carcasses.

Handling time, search time and kill rates

- Handling time went up with prey weight and when the kill was cached, and went down when condors were present. Together these explained 63% of the variation.

- Shorter handling times were followed by shorter search times before the next kill. Prey weight did not affect search time.

- Mean search time was about 70 hours, and pumas moved on average about 20 km between kills (about 33 km with a correction for the real winding path).

- Ungulate kill rates averaged 12.49 kg per day, 50% higher than the 8.28 kg per day reported for pumas in North America. Kill rates did not differ between males and females. Pumas spent on average about 29% of their time at kill sites.

Risk and giving-up time

- GUTs were shortest in the riskiest areas (about 21 to 26 hours in risk classes 6 to 8) and longest in forest (about 94 hours in class 2).

- Risk of condor harassment explained GUTs; prey density did not.

- Pumas made disproportionately many kills in the riskiest areas: 4 to 5 times more often than expected in the two highest risk classes.

- The authors saw condors challenge a puma for a carcass only once. That puma could not hold off attacks from several sides and finally left the kill to the condors. Condors fed in groups of up to 28.

### 3. What they concluded

- Scavengers can impose foraging costs on top predators, and pumas appear aware of the risk of their kills being found by condors in open country.

- Condors shortened the time pumas spent at their kills, and the authors suggest pumas made up for these losses by killing more often.

- Pumas behaved like prey animals in a landscape of fear: they left food sooner where risk was higher, yet hunted most in the high-risk, high-reward open areas where prey was plentiful.

- Caching delayed, but did not stop, condors finding kills in open habitats.

- The authors leave open why pumas abandon kills in the first place: condor harassment, or fear of people, since pumas in open steppe were heavily persecuted by hunters and ranchers. Either way, they argue, condors take the remaining meat, so condors exact the cost.

- Simple single predator, single prey models may not predict real predator and prey dynamics. By raising puma kill rates, scavengers may indirectly affect prey populations and communities.

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

- Finding pumas for capture: the team rode on horseback until they found fresh puma tracks, often in snow, then used hounds to bring the puma to a tree or rock outcrop.

- Kittens: the number of kittens with collared females was found through captures, direct observation, tracks in snow and cameras at kills.

- Kill site investigations: observers identified prey from hair, skin, rumen and bone fragments, and used bite marks and which body parts were eaten to tell a kill from scavenging. They recorded whether carcasses were cached (covered with debris).

- Condor presence: recorded from sightings, cameras, and signs at the kill, including feathers, droppings and footprints. The authors did not try to judge condor presence at kills visited more than two weeks after the puma left, because signs were harder to interpret.

- Walking puma trails: a correction factor for distances travelled came from walking a sample of 2-hour segments of puma trails (described in an earlier paper).

### 5. Tracker or observer certification

Certified trackers. The kill site investigations were done by CyberTracker-certified observers; the paper cites Evans et al. (2009) and Elbroch et al. (2011) for the certification. The level of certification is not given. Neal Wight, Cristián Rivera, Casey McFarland and René Millacúra are thanked for help with kill site investigations.

### 6. Other technologies, skills or innovations

- Argos-GPS collars (SirTrack, Televilt, Lotek, Telonics) and very high frequency (VHF) radio collars, with satellite uplinks every 2 to 5 days.

- Handheld GPS units (Garmin eTrex and Summit) to guide observers to clusters.

- A convex spherical crown densiometer to measure canopy cover over each kill.

- Remote cameras at kills.

- A new measure, giving-up time (GUT), adapted from giving-up density experiments so that wild kills of different sizes could be compared without feeding stations.

- Ages of guanaco and huemul from tooth eruption, and weights from growth estimates, to estimate kilograms killed.

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

- Scavengers usually seen as harmless opportunists may indirectly shape communities by increasing predators’ pressure on their prey.

- Higher puma kill rates mean more encounters with prey, which could be a problem for rare prey. In this area it may partly explain the unsustainable puma predation on endangered huemul.

- The Andean condor is listed as near threatened by the International Union for Conservation of Nature (IUCN).

- Future research should ask whether people or condors drive pumas to abandon kills, and how much condors affect puma survival and breeding.

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

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

- GPS collars showed where and how long pumas stayed, but the questions here were answered on the ground: what was killed, whether it was cached, and whether condors had fed, read from hair, bones, bite marks, feathers, droppings and footprints.

- The authors’ choice not to judge condor presence at sites more than two weeks old is a reminder that signs fade and that sign age limits what can be read at a kill.

- Many sites were visited long after the event (one puma’s sites 650 to 945 days later), which suggests that remains such as bones and hair can still be read for prey long after a kill; the paper does not say how well this worked at those sites.

## Cite this summary

Lawrence, K. 2026. Summary of: Nuisance Ecology: Do Scavenging Condors Exact Foraging Costs on Pumas in Patagonia? (Elbroch and Wittmer, 2013). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2013-elbroch-wittmer-nuisance-ecology-do-scavenging/