Landscape and Urban Planning 94 (2010) 250254

Contents lists available at ScienceDirect

Landscape and Urban Planning

journal homepage: www.elsevier.com/locate/landurbplan

Factors inuencing a motorists ability to detect deer at night

Lauren L. Mastro , Michael R. Conover, S. Nicole Frey
Jack H. Berryman Institute, Department of Wildland Resources, Utah State University, Logan, UT 84322-5230, USA

a r t i c l e

i n f o

Article history:
Received 21 February 2009
Received in revised form
15 September 2009
Accepted 22 October 2009
Available online 26 November 2009
Keywords:
Deervehicle collisions
Roads
Vehicle accidents

a b s t r a c t
Most deervehicle collisions (DVCs) occur at night when deer are active and the ability of motorists
to see them is impaired. The objective of this study was to examine the ability of motorists to detect
deer at night and examine select factors which may inuence detection distances. We examined the
ability of motorists to detect deer at night and the factors that inuence detection using deer decoys,
volunteer motorists, and a series of driving courses. Deer decoys were detected at signicantly greater
distances when located on the right rather than on the left side of the road and when <10 m from the
roads edge. Detection distances were signicantly longer for high-beam headlights versus low-beam
headlights. Moving decoys were detected at similar distances to stationary ones. Additionally, decoys
were detected at greater distances in the absence of reectors than in their presence. The frequency of
nocturnal DVCs can be reduced by using high-beam headlamps and minimizing the number of roadside
reectors, signs, and other bright objects that distract the drivers attention.
2009 Elsevier B.V. All rights reserved.

1. Introduction
Deervehicle collisions (DVCs) are increasing in the United
States and worldwide as trafc volume increases, more roads
are constructed, and habitat becomes more fragmented (Sullivan
and Messmer, 2003). It is estimated that 1.5 million DVCs occur
annually in the United States (Conover, 2001). These collisions
translate annually into >29,000 human injuries, >200 human fatalities, >1.3 million deer fatalities, and >$1 billion in property damage
in the United States (Conover, 1997). Most DVCs (8095%) occur
between sunset and sunrise (Carbaugh et al., 1975; Allen and
McCullough, 1976; Reed and Woodard, 1981; Schafer and Penland,
1985).
While most research has focused on changing deer behavior,
modifying motorist behavior is another way to reduce DVCs. Techniques used to modify motorist behavior include installing road
lighting (Reed and Woodard, 1981) or clearing the side of the road
of vegetation to increase visibility (Lavsund and Sandergren, 1991),
as well as installing deer crossing signs to alert and slow motorists
(Pojar et al., 1975; Reed and Woodard, 1981; Sullivan and Messmer,
2003). These techniques, however, have been largely ineffective.
DVCs are more likely to occur when a motorist fails to see a deer
on the road or sees it too late to stop or take evasive action. At night,
a persons visual abilities, including the ability to recognize obstacles, are reduced as luminance levels decrease and the contrast

Corresponding author. Present address: USDA, APHIS, WS, NWRC, 730 Yokum
Street, Elkins, WV 26241, USA. Tel.: +1 304 636 1785; fax: +1 304 636 5397.
E-mail address: Lauren.L.Mastro@aphis.usda.gov (L.L. Mastro).
0169-2046/$ see front matter 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.landurbplan.2009.10.010

between an object and its background is reduced (Valberg, 2005;

Wood and Owens, 2005). These visual deciencies may cause late
or even failed detection of a deer in night driving situations (Rumar,
1990). In addition, motorists usually look for cars, a logical behavior because cars are the most conspicuous and dangerous objects
on the road. This focus on looking for cars may result in a motorist
failing to see a cyclist, pedestrian, or deer (Rumar, 1990). Studies
show that the average distance at which a pedestrian rst becomes
visible at night on the road is 50 m (Blomberg et al., 1986; Rumar,
1990). This distance is longer when the pedestrian is on the left
side of the road, due to a lack of headlight illumination (SchmidtClausen, 1988; Sivak et al., 1994). Data also show that distances
are also longer when snow is present than when snow is absent
because snow increases contrast with dark objects (Johansson and
Rumar, 1971). Similarly, reective material can be used to increase
the visibility of an object several hundred times by increasing the
objects contrast with its background (Rumar, 1990). To our knowledge, the distance at which a deer is rst visible to a motorist has
not been evaluated.
Motorists abilities to detect deer at night may be increased
through changes in roadside reective posts common to U.S. roadways. The detection and visibility of deer may increase if reectors
are placed at regular intervals so that a deer standing between a
reector and the motorist would block the vehicles headlight from
reecting off one of the posts and back to the motorist. In this situation, a deer standing in or on the edge of the road would create
a gap in the series of reector posts. Motorists may note this and
become aware that something is in the road. Similarly, the ash
(sudden appearance or disappearance) of a reector created as a
deer passes between the motorist and the reector might also make

L.L. Mastro et al. / Landscape and Urban Planning 94 (2010) 250254

251

deer more detectable. If reectors are effective in this manner, they

may reduce the frequency of DVCs.
The objective of this study was to examine the ability of
motorists to detect deer at night and examine select factors which
may inuence detection distances. We tested motorists abilities
to detect deer decoys positioned on the right compared to the left
side of the road and at different distances from the roads edge. We
also studied detection differences between moving and stationary
decoys, low- and high-beam headlamps, as well as the effect of
reector type and spacing.

in which decoys were placed (right, left, and various distances)

along the roadside was randomly assigned, and decoys were placed
>200 m apart. Participants completed the course driving 48 km/h
(30 mph), the speed limit. We used an analysis of variance (ANOVA)
to compare detection distances for decoys on the right and left side
of the road, as well as, 0 and 10 m from the road (Zar, 1999). We
also used an ANOVA to determine detection differences between
decoys placed 0, 10, and 20 m off the right side of the road. For these
and all other tests, results were considered statistically signicant
if P < 0.05.

2. Methods

2.2. Effect of high- and low-beam headlights on detection

distances

We conducted these experiments on sections of two rural roads

(4200 North, 1000 West) in Hyde Park, Utah and at Utah State Universitys Green Canyon Ecology Center in North Logan, Utah. During
these experiments, 4200 North was a 2-lane asphalt road, and 1000
West was a 2-lane graded gravel road. Both roads were bordered by
plowed elds and alfalfa and were in the typical American conguration (left-hand drive vehicles operated on the right side of the
road). The roads at the Green Canyon Ecology Center were 1 lane
and either asphalt or packed clay bordered by low grass. The topography of both sites was at or gently sloping. Each experiment was
conned to a particular road bed type when possible.
Licensed drivers between the ages of 18 and 65+ were recruited
to participate in this study. Each driver only participated once
in each experiment to avoid pseudo-replication but some drivers
participated >1 experiment. The participants drove their personal
vehicles for these experiments because they were familiar with
them. Variation among vehicle windshield cleanliness as well as
headlamp cleanliness and illumination was assumed to accurately
reect actual driving conditions. Participants drove their vehicle
through a driving course with their low-beam headlights on and
alerted the experimenter, sitting in the passenger seat, when they
rst detected 3-dimensional life-sized foam deer decoys (Delta Elite
Mule Deer 3-D Target, Delta Sports Products, LLC, Reinbeck, Iowa).
The experimenter rode with her hand out the passenger side window and dropped a sandbag labeled with the participants assigned
number when the participant indicated he or she saw a decoy.
The distance from the sandbag to the deer decoy was measured
in meters and was dened as the distance at which the participant
rst detected the deer (i.e., detection distance). This distance is less
than the actual distance at which a participant rst detected the
deer decoy because of reaction time and the forward motion of the
vehicle. However, it was determined that this distance was consistent across participants. If a participant failed to detect the decoy
that observation was assigned a value of 0 m. This method whereby
participants are asked to perform a recognition task while seated
in the front seat of a moving car has been performed previously in
studies on the visibility of pedestrians (Luoma et al., 1996).
Experiments were conducted >1 h after sunset on moonless
nights without precipitation or icy roads. Driving courses were congured with various numbers and locations of deer decoys and
reectors. Participants did not know in advance the locations or
numbers of either decoys or reectors along the road. Placement of
decoys along the road was randomized for each experiment and
changed as frequently as possible between participants. Decoys
were always placed perpendicular to the roadway.
2.1. Effect of location on detection distances
We used the 4200 North study site to examine whether the location of a decoy impacted the ability of motorists to detect it. Decoys
were placed 0, 10, and 20 m from the right side and 0 and 10 m from
the left side of the road (0 m signies the edge of the road). The order

To evaluate whether high- or low-beam headlights inuence

detection distances, we placed decoys 0 and 10 m from the right
and left sides of the road. As before, all decoys were spaced >200 m
apart and at randomly assigned locations (right, left, and various
distances) along the roadside at the 4200 North study site. Participants were randomly assigned either high- or low-beams and
completed the course at 48 km/h using the method described previously. The detection distances of decoys 0 m and 10 m from the
road were analyzed separately from each other because of missing
data for some participants. For 10 participants, the decoy located
10 m off the right side of the road was obscured and these data
were not used. For 10 different participants, sandbags were lost for
a decoy 0 m off the right side of the road, so no data were collected.
Therefore, detection distances for decoys at 0 m were analyzed separately from the detection distances of decoys at 10 m. We used an
ANOVA to test for differences in detection distances between headlight strength (high-beam versus low-beam) and side of the road
(right versus left) for decoys at 0 m. We also used an ANOVA to test
for differences in detection distances between headlight strength
and side of the road for decoys at 10 m.
2.3. Effect of reectors on detection distances
To determine if the presence of reectors along the road
inuences the distance at which a motorist rst sees a decoy, participants drove through a driving course at the 4200 North and 1000
West sites. Three sections of road were set up with three types
of reective structures: (1) highway delineator posts that were
orange, free-standing posts measuring 123 cm high and 10 cm in
diameter, with two reectors 7.5 cm high by 10 cm wide, located
80 and 94 cm above the ground, (2) double-reector posts that
were wooden slats measuring 122 cm high by 4 cm wide, with two
reectors 8 cm high by 5 cm wide, located 72 and 108 cm above
the ground, and (3) single-reector posts that were wooden slats
with one reector 8 cm high by 5 cm wide, located 72 cm above
the ground (Fig. 1). The heights of reectors on reector posts were
specically chosen so that a deer passing between the reector and
the motorist (or a decoy placed between the motorist and the reector) would block headlight from reaching the reector. Each type
of reective structure was placed 1 m from the right side of the
road and 40 m apart for a distance of 250 m along the road. Decoys
were then placed 0 m from the right side of the road at randomly
assigned locations for each of the three types of reective structures (i.e., delineator posts, double-reector posts, single-reector
posts). A decoy was also randomly placed in a 250-m section of road
without reectors. Participants completed the course at 48 km/h as
previously described. We used paired t-tests to test for differences
in detection distances between reectors were absent and when
delineator posts, double-reector posts, or single-reector posts
were present (Zar, 1999).
To determine if the distance between reectors along the road
inuences the distance at which a motorist rst detects a decoy,

252

L.L. Mastro et al. / Landscape and Urban Planning 94 (2010) 250254

Table 1
ANOVA table comparing the effect of a decoys location on motorist detection distances of deer decoys at night.
Factor

df

Side of road (right versus left)

Location (0 m, 10 m)
Side location interaction

1,46
1,46
1,46

13.34
0.23
0.01

0.0004
0.64
0.92

side of the decoy opposite the motorist so that they were not visible. Decoys were moved as soon as headlamps were seen entering
that section of the course (>150 m), and the decoys continued to be
moved until the vehicle passed. Participants drove at 24 km/h and
alerted the experimenter as described previously. We used a paired
t-test to examine if detection distances differed between stationary
and moving (1.5 m) decoys when reectors were present. We then
repeated the test without reectors. Results were again analyzed
using a paired t-test.
To determine if there was a difference in detection distances
between when decoys were stationary or moved 5 m, a portion
of this experiment was subsequently repeated. As before, 150m sections of road, one with and one without reectors located
at the Green Canyon Ecology Center were utilized. Each section
had a decoy, which was moved back and forth 5 m, and participants completed the course at 24 km/h as described previously.
We used paired t-tests to examine if detection distances differed
between moving (5 m) and stationary decoys with and without
reectors.

3. Results
Fig. 1. Types of reective structures (from left), single-reector post, doublereector post, and delineator post.

we spaced single-reector posts along a 250-m section of road so

that they were 20 m apart and along another section of road at 40 m
intervals. A deer decoy was randomly placed along each section or
road, and participants completed the course at 48 km/h as previously described. We then used a paired t-test to compare detection
distances for decoys with reectors spaced 20 and 40 m apart (Zar,
1999).
To evaluate detection differences when the number of reectors on a single post differs, we used two types of delineator posts
on two sections of road at the Green Canyon Ecology Center. Delineator posts with two reective strips (7.5 cm high by 10 cm wide, at
heights of 81 and 95 cm) and with four reective strips (at heights
of 54, 67, 81, and 95 cm) were placed 1 m from the right side of
the road and 20 m apart for 150 m. Participants drove at 24 km/h
(15 mph) alerting the experimenter as previously described. A
paired t-test was used to determine differences in detection distances between delineator posts with two and four reective strips
(Zar, 1999).

3.1. Effect of location on detection distances

There were longer detection distances when decoys were on the
right side (x = 43 m, SE = 4, n = 48) than on the left side (x = 23 m,
SE = 3, n = 48) of the road (Table 1). Detection distances did not differ
between decoys located 0 m (x = 34 m, SE = 4, n = 48) and 10 m (x =
31 m, SE = 5, n = 48) off the road (Table 1). However, decoys 20 m
from the right side of the road were detected at signicantly shorter
distance than those placed 0 or 10 m from the right side of the road
(F = 42.66, df = 2,23, P < 0.0001, Fig. 2). Only 19% of 26 participants
observed decoys located 20 m off the right side of the road while
all participants detected a decoy 0 or 10 m off the right side of the
road.

2.4. Impact of deer movement on detection distances

To examine detection differences when decoys were stationary
versus when they were moving, we used two 150-m sections of
road at the Green Canyon Ecology Center. Each section had eight
reector posts placed 20 m apart and 1 m from the right side of
the road. Either a stationary or moving decoy was positioned 0 m
from the right side of the road at randomly assigned locations in
each of the two sections. Decoys were moved 1.5 m back and forth
perpendicular to the road, so that the movement of the decoy mimicked the movement of a deer entering the roadway. Volunteers
moving the decoys, wore dark clothing and were positioned on the

Fig. 2. Mean detection distances of decoys located at 0, 10 and 20 m off the right
side and 0 and 10 m off the left side of the road at night (the bars show the standard
error of each mean).

L.L. Mastro et al. / Landscape and Urban Planning 94 (2010) 250254

Table 2
ANOVA table comparing motorist detection distances of deer decoys at night with
high and low-beam headlights and for different sides of the road (decoys were
positioned 0 m from the road).
Factor

df

Headlight beam (high versus low)

Side of road (right versus left)
Beam side interaction

1,12
1,12
1,12

7.01
6.00
0.01

0.01
0.02
0.94

3.2. Effect of high- and low-beam headlights on detection

distances
Detection distances were shorter for low beams (x = 41 m,
SE = 6, n = 14) than high beams (x = 73 m, SE = 11, n = 14) when
decoys were located 0 m from the road (Table 2). Detection distances were greater for decoys 0 m from the right side (x = 69 m,
SE = 8, n = 14) than for decoys 0 m from the left side (x = 45 m,
SE = 10, n = 14) of the road (Table 2). Detection distances for decoys
10 m off the road did not differ between high and low-beam headlights (F = 1.94, df = 1,12, P = 0.15).
3.3. Effect of reectors on detection distances
Detection distances were shorter when either delineator posts
or double-reector posts were present than when reective structures were absent (Table 3). However, there was no difference
in detection distances when single-reector posts were present
versus when reective structures were absent (Table 3). Placing
reector posts closer together (20 m apart versus 40 m) decreased
detection distances (t = 3.06, df = 14, P = 0.008). Detection distances did not differ signicantly between delineator posts with
two versus four reectors (t = 0.62, df = 10, P = 0.55).
3.4. Impact of deer movement on detection distances
There were no signicant differences in detection distances
between stationary decoys and decoys that moved 1.5 m (t = 0.48,
df = 9, P = 0.64) or 5 m (t = 1.35, df = 19, P = 0.19) in the presence of
reectors. Likewise, when reectors were absent, stationary decoys
were no more conspicuous than decoys that moved 1.5 (t = 1.45,
df = 9, P = 0.18) or 5 m (t = 1.93, df = 19, P = 0.07).
4. Discussion
In our study, all motorists were aware that decoys would be
present somewhere along the driving course but did not know their
exact location or number. Hence, the detection distances we found
are probably greater than would occur in an actual driving situation when motorists would be unaware that deer are present and
therefore even less likely to detect them because they are not looking for them. Still, most motorists in our tests failed to see a deer
standing on the roads edge until they were within 50 m of it. At
this distance, motorists would not have been able to stop in time
to avoid hitting a deer if their vehicle was traveling at >75 km/h.
When decoys were located 20 m off the road, most motorists (81%)
Table 3
Results of t-tests comparing motorist detection distances of deer decoys at night
when different types of reectors were lining the road versus when no reectors
were present.
Type of reector

SE

df

None
Single-reector post
Double-reector post
Delineator post

85
84
59
45

9
9
5
5

15
15
15
15

0.09
3.33
5.39

14
14
14

0.93
0.005
<0.0001

253

in our study never saw them. Our results show that deer located a
short distance from a road are unlikely to be detected at night and
can pose a threat to drivers when they dart across the road in front
of a vehicle.
We found detection distances of decoys were similar to those
reported for pedestrians. Decoys located 0 m from the right side
of the road without snow were detected at an average of 48 m
while pedestrians under similar conditions were detected at 50 m
(Blomberg et al., 1986; Rumar, 1990). Detection distances of decoys
were greater on the right side versus the left side of the road which
was consistent with reported detection distances of pedestrians
(Schmidt-Clausen, 1988; Sivak et al., 1994).
As expected, high-beam headlights were associated with longer
detection distances than low-beam headlights. High-beam headlights increased the distance at which a motorist saw a decoy by
21 m when the decoy was on the right side of the road and 37 m
when the deer was on the left side.
We expected that a conspicuous gap in the series of reective structures along the road created by the presence of a decoy
blocking a vehicles headlight from reecting off the post and back
to the motorist would increase detection distances. However, the
opposite was true. The presence of reectors decreased detection
distances. Similarly, shortening the distance between reectors
(from 40 to 20 m) so that decoys were more likely to create gaps
in the series of reectors decreased detection distances. This may
be because participants were focused on the reectors and failed
to detect the less conspicuous deer decoys.
Surprisingly, moving decoys were not easier for a motorist to
detect than stationary ones. We had expected moving decoys to be
easier to detect than stationary ones especially in the presence of
reectors. In this situation, moving decoys caused a ash (sudden
appearance or disappearance) of a reector as they passed between
the motorist and the reector, alternately blocking and unblocking headlight from reaching the reector and reecting back to the
motorist. We expected this ash to increase detection distances but
it did not do so.
We found that few motorists could detect a deer decoy standing beside the road at night unless it was <10 m from the side of
the road. Even then, most motorists did not see the deer decoy
until they were within 50 m of it. Detection distances increased
when motorists used their high-beam headlights. Hence, motorists
should be encouraged to use them in areas where DVCs are common. At night, our results suggest that a motorists attention is
drawn to items on or near the road that are more visually conspicuous than deer (e.g., reectors, dark objects against a snowy
background). Distractions delay a motorists ability to see deer.
Therefore, anything that minimizes distractions should increase
deer detection distances and may reduce DVCs.
Acknowledgments
The Jack H. Berryman Institute and the Utah Agricultural Experiment Station provided funding for this study. We thank numerous
study participants, without whom this study would not have been
possible. Special thanks to J. Borgo and K. Sivey for their help conducting these tests.
References
Allen, R.E., McCullough, D.R., 1976. Deer-car accidents in southern Michigan. J.
Wildlife Manage. 40, 317325.
Blomberg, R.D., Hale, A., Preusser, D.F., 1986. Experimental evaluation of alternative
conspicuity-enhancement techniques for pedestrians and bicyclists. J. Safety
Res. 17, 112.
Carbaugh, B., Vaughan, J.P., Bellis, E.D., Graves, H.B., 1975. Distribution and activity of
white-tailed deer along an interstate highway. J. Wildlife Manage. 39, 570581.
Conover, M.R., 1997. Monetary and intangible valuation of deer in the United States.
Wildlife Soc. B 25, 298305.

254

L.L. Mastro et al. / Landscape and Urban Planning 94 (2010) 250254

Conover, M.R., 2001. Resolving HumanWildlife Conicts. CRC Press, Boca Raton, FL.
Johansson, G., Rumar, K., 1971. Silhouette effects in night driving. Scand. J. Psychol.
12, 8089.
Lavsund, S., Sandergren, F., 1991. Moose-vehicle relations in Sweden: a review. Alces
27, 118126.
Luoma, J., Schumann, J., Traube, E.C., 1996. Effects of retroreector positioning on
nighttime recognition of pedestrians. Accid. Anal. Prev. 28, 377383.
Pojar, T.P., Prosence, R.A., Reed, D.F., Woodward, T.N., 1975. Effectiveness of a lighted,
animated deer crossing sign. J. Wildlife Manage. 39, 8791.
Reed, D.F., Woodard, T.N., 1981. Effectiveness of highway lighting in reducing
deervehicle accidents. J. Wildlife Manage. 45, 721726.
Rumar, K., 1990. Basic driver error: late detection. Ergonomics 33, 12811290.
Schafer, J.A., Penland, S.T., 1985. Effectiveness of Swareex reectors in reducing
deervehicle accidents. J. Wildlife Manage. 49, 774776.
Schmidt-Clausen, H.J., 1988. Recognition of pedestrians during nighttime trafc. In:
Proceedings of the 11th International Technical Conference on Experimental
Safety Vehicles. U.S. Department of Transportation and National Trafc Safety
Administration, Washington, DC, pp. 456459.
Sivak, M., Flannagan, M.J., Sato, T., 1994. Light output of U.S., European, and Japanese
low-beam headlamps. Transport. Res. Rec. 1456, 99111.

Sullivan, T.L., Messmer, T.A., 2003. Perceptions of deervehicle collision management by state wildlife agency and department of transportation administrators.
Wildlife Soc. B 32, 907915.
Valberg, A., 2005. Light, Vision, Color. John Wiley and Sons, West Sussex, England.
Wood, J.M., Owens, D.A., 2005. Standard measures of visual acuity do not predict
drivers recognition performance under day or night conditions. Optom. Vis. Sci.
82, 698705.
Zar, J.H., 1999. Biostatistical Analysis, Fourth edition. Prentice Hall, Upper Saddle
River, NJ.
Lauren L. Mastro received her MS in wildlife biology from Utah State University and
her BS degree in biology and environmental science from Muhlenberg College. She
is currently employed by USDA-APHIS-Wildlife Services-National Wildlife Research
Center. She is a member of the Jack H. Berryman Institute, the Wildlife Society, and
the Wildlife-Damage Working Group.
Michael R. Conover is a faculty member with the Jack H. Berryman Institute and
the Department of Wildland Resources, Utah State University.
S. Nicole Frey is a faculty member of the Jack H. Berryman Institute and Southern
Utah University.