Conservation Evidence strives to be as useful to conservationists as possible. Please take our survey to help the team improve our resource.

Providing evidence to improve practice

Action: Scare or otherwise deter mammals from human-occupied areas to reduce human-wildlife conflict Terrestrial Mammal Conservation

Key messages

Read our guidance on Key messages before continuing

  • Ten studies evaluated the effects of scaring or otherwise deterring mammals from residential areas to reduce human-wildlife conflict. Six studies were in the USA, three were in Canada and one was in Tanzania.





  • Human-wildlife conflict (10 studies): Two of four studies (including one randomized and controlled study) in the USA, found that a range of noise and pain deterrents did not prevent black bears from returning to urban areas or other human-occupied sites. The other two studies found that such actions did deter them from seeking food at human-occupied sites. Two of three studies, in the USA and Canada, found that chasing nuisance black bears with dogs and chasing elk with people or dogs caused them to stay away longer or remain further from human occupied areas. The other study found that attempts to scare coyotes did not cause them to avoid human occupied areas. A before-and-after study in Canada found that an electric fence prevented polar bear entry to a compound. A study in Canada found that chemical and acoustic repellents did not deter polar bears from baits in most cases. A replicated study in Tanzania found that drones caused African savanna elephants to quickly leave residential areas.

Supporting evidence from individual studies


A before-and-after study in 1983–1985 at a research compound in Manitoba, Canada (Davies & Rockwell 1986) found that after the area was enclosed with an electric fence, no polar bears Ursus maritimus entered it. Over a total of approximately five months over two summers with the fence installed, no polar bears entered the compound. However, before the fence was installed in those years and in the previous year before it was first installed, nine different bears visited the compound, some on multiple occasions. The study was conducted in a research compound where 10–15 biologists resided between May and September each year. In July–September 1984 and June–September 1985, a temporary two-strand electric fence was erected around the 300-m compound perimeter. The two strands of wire were 30 and 60 cm above the water or ground. The fence emitted 40 pulses/min of direct current (peak output of 8,000 volts). When the fence activated, two 110-decibel horns also sounded.


A study in 1978 at a shrubland and grassland site in Manitoba, Canada (Miller 1987) found that acoustic deterrents and baits treated with chemical deterrents did not, in most cases, repel polar bears Ursus maritimus. Out of 55 visits, acoustic deterrents repelled bears on 17 visits and did not repel them on 38 visits. From 294 visits, chemical deterrent repelled bears five times but did not repel them during 289 visits. However, bears remained for shorter periods at chemical repellent-treated bait stations (average 98–317 s) than at baits without repellents (average 420 s). In October–November 1978, polar bears were attracted to 13 bait stations with sardines. Stations were all 100–500 m from a 6-m-high tower, from which bear responses were observed. At one bait station, a loudspeaker was placed 5 m from the bait. Sounds played through the loudspeaker included bear sounds, human shouting, killer whale sounds, radio noise and human hissing and barking like a bear. Ten bait stations were sprayed with dog-repellents or household chemicals. Two bait stations had no repellents.


A study in 1990–1998 of a largely forested national park in North Carolina and Tennessee, USA (Clark et al. 2002) found that following capture and release back at capture sites, most black bears Ursus americanus did not subsequently repeat nuisance behaviour, such as entering picnic sites or campgrounds. For 50 out of 85 captures, bears were not subsequently sighted at capture locations during the remainder of that year. In four further cases, no management action was required that year, even if the bear was resighted at its capture location. In a 2,080-km2 national park, 63 bears exhibiting nuisance behaviour (such as raiding bins) were captured by live-trapping or darting. Bears were immobilised, individually marked and had a tooth extracted (for aging) before release, after recovery from anaesthesia, <150 m from their capture site.


A randomized, controlled study in 1997–2002 in residential areas and adjacent forest across at least four mountain ranges in Nevada, USA (Beckmann et al. 2004) found that subjecting nuisance black bears Ursus americanus to deterrents intended to scare them, did not prevent their return to urban areas. The average time for bears to return to urban areas after treatments did not differ significantly between those chased by dogs Canis lupus familiaris in addition to noise and projectile deterrents (154 days), those subject to the same deterrents excluding chasing by dogs (88 days) or those not subject to deterrents (65 days). Fifty-seven of the 62 bears in the study returned to urban areas. Forty-four of these returned within 40 days. Nuisance bears (which raided garbage) were captured and radio-collared between July 1997 and April 2002. They were randomly assigned to deterrent treatments including chasing by dogs (20 bears), deterrent treatments excluding chasing by dogs (21 bears) or no deterrent (20 bears). Additional to chasing by dogs, deterrents entailed pepper spraying, firing 12-gauge rubber buckshot or rubber slugs, loud cracker shells and shouting. Deterrents were administered at release sites, 1–75 km from capture locations.


A replicated, controlled study in 2004 of ten forest sites in Minnesota, USA (Breck et al. 2006) found that installing electric shock devices prevented American black bears Ursus americanus from accessing or damaging bird feeders. Bird feeders protected by electric shock devices suffered less bear damage (none of ten accessed or damaged) than did unprotected feeders (four of ten accessed or destroyed). Two imitation bird feeders were installed at each of ten sites, ≥30 km apart. One feeder was protected by an electric shock device, the Nuisance Bear Controller. This device had two 6-volt batteries wired to an automobile vibrator coil/condenser, emitting 10,000–13,000 volts through a disk when contact is made by an animal. The other feeder was unprotected. Ground around each feeder was cleared to enable identification of bear signs. Feeders were in place from 1 July to 15 November 2004. They were monitored, and bait replenished, at least weekly.


A controlled study in 2001–2002 at a town and surrounding forest in Alberta, Canada (Kloppers et al. 2005) found that after being chased by humans, the average distance of elk Cervus canadensis from the town increased more than it did for elk chased by dogs Canis lupus familiaris or for elk that were not chased. The average distance of elk from the town boundary increased for all treatment groups but the increase was larger for elk chased by humans (after: 1,130 m; before: 184 m) than for elk chased by dogs (after: 1,041 m; before: 535 m) or for elk that were not chased (after: 881 m; before: 629 m). Twenty-four elk were radio-collared. Each was assigned to being chased by humans, chased by dogs or not chased, 10 times, from November 2001 to March 2002. Chases lasted 15 minutes and covered averages of 1,148 m when humans (shooting starter pistols) chased elk and 1,219 m when two border collie dogs chased elk. Non-chased elk moved an average of 49 m during 15 minutes. Capture and collar-fitting may have produced some aversive response though animal handling was uniform across groups. Displacement from the town boundary was calculated from daily sightings or radio-signals, from September 2001 to March 2002.


A study in 2005–2006 at a site comprising marsh, forest, farmland, and residential areas in Louisiana, USA (Leigh & Chamberlain 2008) found that chasing nuisance black bears Ursus americanus with dogs Canis lupus familiaris, in addition to making noise and shooting with rubber buckshot, increased the amount of time until they next exhibited nuisance behaviour compared to solely making noise and shooting rubber buckshot. Black bears subjected to chasing by dogs, loud noise and shooting with rubber buckshot took longer to return to nuisance behaviour (58 days) than did bears that were subjected to loud noise and shooting with rubber buckshot but not chasing by dogs (48 days). Between April 2005 and July 2006, eleven bears reported to be exhibiting nuisance behaviour were live-trapped. All were immobilized and fitted with radio-collars. Upon release, six bears were subjected to loud noise, shooting with rubber buckshot and chasing with dogs and five were subjected to loud noise and shooting with rubber buckshot alone. Bears were monitored for recurring nuisance behaviour for up to 5 months after release.


A study in 2002–2005 in a national park in California, USA (Mazur 2010) found that aversive conditioning reduced the number of black bears Ursus americanus that were accustomed to seeking food at human-frequented locations revisting. Of 29 bears accustomed to taking human-food, 17 ceased to do so, six required continued aversion conditioning and six “persistent offenders” were removed or killed for safety reasons. Over 150 bears were subject to 1,050 aversive conditioning events. Of these, 729 events involved 36 individual food-conditioned or habituated bears (seven became habituated in the final year of the study, so their subsequent behaviour was not assessed). Five personnel drove bears from campsites and other human-occupied areas by throwing rocks and using sling shots, pepper spray, rubber slug projectiles and chasing. All actions were accompanied by shouting. Aversive conditioning actions were carried out each summer, from June 2002 to September 2005.


A replicated, controlled study in 2014 of four urban areas in Colorado, USA (Breck et al. 2017) found that attempts to scare away coyotes Canis latrans did not decrease their use of areas also frequently used by people. On trails frequently travelled by people, the overlap between coyote and human activity was similar where community-level programmes were run to scare coyotes and where programmes were not run (data presented as coefficients of overlap, incorporating frequency and timing of use). On trails with less human traffic, overlap between coyote and human activity was greater where programmes were run than where they were not run. These differences were not tested for statistical significance. Four urban park and open space areas were studied. In two, community-level programmes were run. These primarily involved shouting, throwing objects, and/or aggressively approaching coyotes. Activities were promoted by signs, social media, emailing to multiple recipients, education stations and an online video. Programmes were not run in the two control areas. Coyote and human use of trails were monitored using five camera traps in each area for a 3–4-week period, generating >50,000 independent records of people and coyotes.


A replicated study in 2016 in two savanna reserves in Tanzania (Hahn et al. 2017) found that using drones to deter African savanna elephants Loxodonta africana from towns led to elephants leaving the sites quickly. On all 13 occasions, when drones were deployed, elephants began to flee within one minute. Elephants were typically herded to an area > 1 km from villages. Before using drones, rangers were trained during three 4-day workshops. In February–March and May–August 2015 and in March–April 2016, rangers deployed drones in 13 situations when elephants were found close to villages. Each drone was fitted with a flashlight, to locate elephants at night, and, during the day, a live video feed from a camera on the drone was used. Elephant responses were recorded over 60-second intervals for the first 10 minutes of the drone flight.

Referenced papers

Please cite as:

Littlewood, N.A., Rocha, R., Smith, R.K., Martin, P.A., Lockhart, S.L., Schoonover, R.F., Wilman, E., Bladon, A.J., Sainsbury, K.A., Pimm S. and Sutherland, W.J. (2020) Terrestrial Mammal Conservation: Global Evidence for the Effects of Interventions for terrestrial mammals excluding bats and primates. Synopses of Conservation Evidence Series. University of Cambridge, Cambridge, UK.