Fit vehicles with ultrasonic warning devices
Overall effectiveness category Unlikely to be beneficial
Number of studies: 3
Background information and definitions
Collisions between mammals such as deer and vehicles can result in death or injury to animals and humans alike. For example, it has been estimated that over 1 million deer-vehicle collisions occur annually in the USA alone (Conover et al. 1995). Wildlife warning whistles are designed to produce high frequency, ultrasonic noises to alert or frighten animals away from oncoming vehicles. Whistles can be mounted on vehicles, with the sound being emitted once the vehicle reaches a certain speed. Alternatively whistles can be mounted on poles or small trees along roads and be activated by headlights of approaching cars.
Conover M.R., Pitt W.C., Kessler K.K., DuBow T.J. & Sanborn W.A. (1995) Review of human injuries, illnesses, and economic losses caused by wildlife in the United States. Wildlife Society Bulletin, 23, 407–414.
Supporting evidence from individual studies
A controlled study in 1990 in sagebrush in Utah, USA (Romin & Dalton 1992) found that vehicle mounted wildlife warning whistles had no effect on the behaviour of mule deer Odocoileus hemionus. The proportions of deer that responded to the vehicle were 31% with a whistle and 39% without. Six percent of deer ran away from the vehicle with a whistle and 12% did so from the vehicle without a whistle. Authors reported that they did not know if the whistles produced any sound, nor if deer heard them. Two brands of wildlife warning whistles (Game Tracker's and Sav-a-life, producing 16–20 kHz) were mounted on the front of a truck. These were tested during late afternoon and early evening along 9.7 km of dirt road in January–February 1990. For each of 150 groups of deer (average six deer), a pass at 65 km/hour was made without and then with the whistle. Deer responses (none, head lifted, changed orientation, ran away, ran towards) and distances from the road were recorded for each pass (distances did not differ significantly between first and second passes).Study and other actions tested
A replicated, controlled study in 1997–2001 along roads in New South Wales, Queensland, Victoria and Western Australia, Australia (Bender 2001) found that Shu Roo warning whistles did not alter behaviour of eastern grey kangaroos Macropus giganteus or red kangaroos Macropus rufus and did not reduce kangaroo-vehicle collisions. There was no significant difference in the number of kangaroos hit by vehicles with or without whistles (22% with; 7% without). Vigilance responses did not differ significantly for either species when whistles were turned on (60–65%) or off (40–75%) and no animals fled in response. The Shu Roo was not purely ultrasonic (4–19 kHz) and was only detected at 50 m. The whistle was not detectable above the noise of the four vehicles tested. The Shu Roo (two speakers in a rectangular metal case) signal was tested in the lab and in the field at 20–400 m (static and mounted on four vehicle types). Responses of 31 captive kangaroos to the Shu Roo (turned on/off), mounted on a vehicle at 20–50 m, was recorded on 15 occasions in July–September 1997. Fifteen companies, in which people travelled large distances (average 49,000 km) conducted surveys in four states in August 1999 to January 2001. Fifty-seven vehicles had a Shu Roo fitted and 40 vehicles did not.Study and other actions tested
A replicated, controlled study in 2006 at a college campus in Georgia, USA (Valitzski et al. 2009) found that high frequency sounds from moving vehicles did not reduce white-tailed deer Odocoileus virginianus behaviours that were likely to cause a deer–vehicle collision. At 0.28 kHz, there was a significant increase in the proportion of behaviours likely to cause a collision (13%) compared to a vehicle without treatment (5%). At four other frequencies, there was no significant difference in proportions of negative behavioural responses compared to the vehicle without treatment (1–28 kHz: 6–9%). The proportion of behaviours likely to decrease deer-vehicle collisions did not differ between different high frequencies and no high-frequency sound (0.28 kHz: 33%; 1 kHz: 37%; 8 kHz: 24%; 15 kHz: 33%; 28 kHz: 24%; no high-frequency sound: 35%;). Two road sections (≥ 5 km apart), 280 m and 220 m long, were studied. For each of 319 trials, a deer was observed before and during one of six randomly assigned treatments: 0.28, 1, 8, 15 or 28 kHz or no sound. The high-frequency sounds (within deer hearing range) were played at 70 decibels from front-mounted speakers on the vehicle (48 km/hr). Deer within 10 m of the road or ahead of the vehicle were monitored from an observation platform, from 06:00 to 09:00 h and 19:00 to 22:00 h, in April and June 2006.Study and other actions tested