Wildlife Society Bulletin 37(2):402–408; 2013; DOI: 10.1002/wsb.253
Tools and Technology
Effectiveness of Wildlife Guards at Access Roads
1
MARCEL P. HUIJSER, Western Transportation Institute, Montana State University, P.O. Box 174250, Bozeman, MT 59717, USA
DAVID W. WILLEY, Department of Ecology, Montana State University, 310 Lewis Hall, Bozeman, MT 59717, USA
ABSTRACT The reconstruction of 90.6 km of U.S. Highway 93 from Evaro to Polson, Montana, USA,
includes 41 wildlife crossing structures and 13.4 km of road with wildlife fencing. These measures are
aimed at reducing wildlife–vehicle collisions and increasing human safety, while allowing wildlife to
traverse the landscape. In the fenced road sections, gaps in the fence for side roads are mitigated by wildlife
guards (similar to cattle guards). We monitored wildlife movements with cameras for 2 years from midJuly 2008 to mid-July 2010 at 2 wildlife guards and in 1 large crossing structure adjacent to one of the
wildlife guards. We investigated how effective these wildlife guards were as a barrier to deer (Odocoileus
spp.), black bear (Ursus americanus), and coyotes (Canis latrans). We also compared movements across a
wildlife guard with movements through an adjacent crossing structure. The wildlife guards were �85%
effective in keeping deer from accessing the road and 93.5% of deer used the crossing structure instead of
the adjacent wildlife guard when crossing the road. The wildlife guards were less effective in keeping black
bear and coyotes from accessing the road (33–55%). However, all black bears and 94.7% of coyotes used
the crossing structure instead of the adjacent wildlife guard when crossing the road. Though the wildlife
guards were not an absolute barrier to these species, the results indicate wildlife guards are a substantial
barrier to deer and can be considered effective in mitigating gaps in a fence at access roads for these
species. ß 2013 The Wildlife Society
KEY WORDS deer, fence, mitigation, Montana, Odocoileus spp., road ecology, wildlife guard.
There are currently >8.4 million lane-miles of highway in the United States, with >6.1 million lane-miles in
rural areas (Federal Highway Administration 2007). The ecological effects of roads and vehicles are
diverse and include 1) loss of habitat due to pavement or other unnatural substrate; 2) direct mortality by
collisions with vehicles; 3) habitat fragmentation due to barriers that affect animal movements; and 4)
reduced habitat quality adjacent to roads (e.g., because of chemical or noise pollution; Forman and
Alexander 1998, Beckmann et al. 2010).
Roads not only affect wildlife, but people are also at risk when large mammals enter the roadway. Between
1 and 2 million collisions with large animals occur in the United States each year, with about 29,000 human
injuries and 200 human deaths (Conover et al. 1995, Huijser et al. 2008). Total estimated costs of animal–
vehicle collisions exceed US$ 8 billion annually, including costs associated with vehi- cle repair, human
injuries and fatalities, accident investiga- tion, carcass removal and disposal, and the monetary value of the
animal to hunters (Huijser et al. 2009). Furthermore, �42 mitigation measures or combinations of
mitigation measures have been implemented or suggested to mitigate
Received: 28 June 2011; Accepted: 24 September 2012 Published: 18 March 2013
1
impacts, most of which have not been thoroughly studied. These alternatives range from public information
to wildlife fencing and from roadside animal-detection systems to cull- ing wildlife (Farrell et al. 2002,
Huijser et al. 2008).
Wildlife fencing can substantially reduce ungulate access to the road corridor (Falk et al. 1978) and has
been shown to reduce wildlife–vehicle collisions by �79% (Reed et al. 1982, Clevenger et al. 2001, Dodd
et al. 2007). However, in developed areas, gaps in fences are necessary to allow vehicles access to and from
main roadways. Without additional measures, gaps can allow wildlife to enter the fenced road corridor.
Though gates (locked or unlocked) have been used at low-use side roads, they are sometimes left open and
are not suitable for higher traffic volumes. In contrast, wildlife guards, devices that are similar to cattle
guards, appear to be an interesting alternative because they are not only designed to handle relatively high
traffic volumes but can also discour- age wildlife from entering the fenced right-of-way at gaps in the fence
for access roads (e.g., Peterson et al. 2003).
Traditional and modified cattle guards have been found effective for white-tailed deer (Odocoileus
virginianus) under some circumstances (Bashore and Bellis 1982, Belant et al. 1998, VerCauteren et al.
2009). However, to be suitable for public roads and effectively exclude wildlife from the fenced road
corridor, several factors must be considered. Standard cattle guards may not be safe for pedestrians,
cyclists, and
�
Figure 1. Locations of 2 wildlife guards and 5 crossing structures in the 1-km study area along U.S. Highway 93, south of Ravalli,
Montana, USA.
heavy vehicles so alternative designs may be required (Peterson et al. 2003). In addition, side panels
(fencing) perpendicular to the main highway should be positioned along the full length of the wildlife guard
to prevent diagonal (shorter) crossing by wildlife (Sebesta et al. 2003). Wildlife guards should also be
placed over a pit with walls (e.g., concrete) or raised in another manner to prevent wildlife from crossing by
placing their feet on the ground between the grates (Sebesta et al. 2003).
In this field study, our objective was to investigate how effective a particular type of wildlife guard design
was as a barrier to mule deer (O. hemionus) and white-tailed deer movement and, therefore, how effective
it was in keeping deer from accessing the right-of-way along a fenced section of U.S. Highway 93. We also
opportunistically investigated the effectiveness of the wildlife guard design as a barrier to other medium to
large mammal species for which we had a sample size of �10: black bear (Ursus americanus) and coyotes
(Canis latrans).
STUDY AREA
From October 2004 to November 2010, there were 8 recon- struction projects along 90.6 km of U.S.
Highway 93 from Evaro to Polson, located on the Flathead Indian Reservation in northwestern Montana,
USA. These projects were intended to alleviate congestion and improve safety and included wider
shoulders, passing lanes, turning lanes, im- proved sight-distances, and wildlife mitigation measures. The
project corridor was the site of one of the most extensive wildlife crossing mitigation projects in North
America, both in terms of road length and the number of wildlife crossing structures. There were 41
structures, excluding likely future
mitigation measures in the Ninepipes wetland area. In addi- tion to crossing structures that allow animals to
cross under or over the road safely, the reconstruction included 2.4-m- high wildlife exclusion fencing
along some road sections intended to restrict access of wildlife to the road corridor.
The Flathead Indian Reservation is home to the Confederated Salish and Kootenai Tribes, which includes
the Bitterroot Salish, Pend d’Oreille, and Kootenai tribes. Our study was conducted along a section of U.S.
Highway 93, which ran north–south through this reservation. This section, located just south of Ravalli and
the junction with Montana Highway 200, was 1 km long and was referred to as ‘‘Ravalli Curves’’ (Fig. 1).
There were several mitigation measures in our study area, including wildlife fencing, wild- life
underpasses, wildlife guards, and jump-outs (earthen ramps that allowed animals that may enter the fenced
road corridor to jump down to safety; see e.g., Bissonette and Hammer 2000, Huijser et al. 2009 for further
details). The wildlife underpasses varied in type and size. There were 3 medium mammal culverts, 1 large
mammal underpass, and 1 bridge (Table 1). There were also 2 wildlife guards and 9 jump-outs in the study
area.
Table 1. Dimensions (m) of wildlife crossing structures along U.S. Highway 93, south of Ravalli, Montana, USA.
Underpass type
Culvert Culvert Culvert Culvert Bridge
Height Width Length
3.5 7.7 18.4 1.2 1.8 25.0 1.4 2.8 21.7 1.6 1.2 21.3 3.9 28.4 11.5
�
403
Both wildlife guards were located on the east side of U.S. Highway 93 and had 2.4-m-high fencing on each
side of the wildlife guard, perpendicular to U.S. Highway 93. Each wildlife guard consisted of a steel grate
placed over a back- filled depression supported by concrete foundation walls. The grates were 6.8 m wide
in the direction parallel to the traffic on U.S. Highway 93 by 6.6 m long in the direction perpendicular to
traffic. Each grate was formed by smaller 8-cm � 10-cm rectangles, made of a combination of 6-mm and
130-mm steel. The depressions were approximately 76 cm deep and were filled with approximately 46 cm
of backfill, effectively leaving about a 45-cm-deep pit after the steel grate was placed on top. This pit
served to discourage animals from placing their feet on the ground between the grates. Foundation walls of
the depression were 20 cm wide, surrounded the metal grate on all sides, and both were manufactured by
L.B. Foster Company (Pittsburgh, PA, USA; Fig. 2).
The landscape adjacent to this 1.0-km section of mitigation measures was relatively homogeneous, with a
railroad and the Jocko River running parallel to U.S. Highway 93 on the west side, an old oxbow of the
Jocko River to the east, and steep hills to both the east and west. The surrounding vegetation was primarily
coniferous forest with some grass- land, as well as riparian habitat along the river and old oxbow. The
average daily traffic in our study area was 7,205 during the pre-construction years 1999–2000 (Montana
Department of Transportation [MDT] 2008). The average daily traffic during post-construction monitoring
decreased by 4.6% to 6,872 (MDT 2008, 2009, 2010).
METHODS
We evaluated the effectiveness of the wildlife guards as a barrier to wildlife movements by addressing 3
research ques- tions. First, how effective are the wildlife guards as a barrier to animals that come close to
the wildlife guard or that show intent to cross? To answer this question, we calculated the percentage of
animal movements that crossed the wildlife guards out of the number of animal movements that
approached the wildlife guards traveling toward the fenced road corridor. Second, are wildlife guards a
greater barrier to animals than the road and traffic alone? We answered this question by comparing preconstruction crossing rates at the road with crossing rates at the wildlife guard, post-construc- tion. Third,
we asked, if presented with an option, will animals use a wildlife guard or a crossing structure to cross the
road? To answer this question, we compared animal movements across one of the wildlife guards with
those through a large crossing structure (about 7.7 m wide, 3.5 m high) near the wildlife guard, 61 m south.
We monitored wildlife movements for 2 years from mid-July 2008 to mid-July 2010 with wildlife cameras
TM
the 2 wildlife guards and one in the culvert. In our analyses, we included all wildlife species for which the
sample size was >10 groups. We included only east to west movements in the analyses because east to west
movements at the wildlife guards were likely motivated by intent to cross the road rather than to escape the
fenced road corridor. For compari- son purposes, we also selected east to west movements for the culvert.
In addition, for all movements at the 2 wildlife guards and the culvert, we counted animals of the same
species that were traveling together as one independent animal record, because their movements were likely
interdependent.
We calculated the barrier effect of the wildlife guards in 2 different ways depending on whether an animal
movement fit one or both of the following definitions: 1) the animal came within 2 m of the wildlife guard;
2) the animal dis- played behavior indicating it intended to cross the wildlife guard. For the second
definition, qualifying behavior includ- ed, but was not limited to, stalling at the wildlife guard, placing its
nose on the wildlife guard, and pawing at or stepping onto the wildlife guard. Though this was a subjective measure, one observer analyzed and categorized all images to ensure consistency. We used both a
distance measurement and a behavior measurement because there
Figure 2. One of 2 identical wildlife guards and dimensions of the wildlife guards (A) and grates (B) on access roads to U.S. Highway
93, south of Ravalli, Montana, USA.
�
were strengths and weaknesses to both methods. The dis- tance measurement was objective. However, an
animal may have passed a wildlife guard within 2 m (distance measure- ment) without being interested in
crossing, potentially resulting in an overestimation of the barrier effect. The behavior measurement left no
question that the animal noticed the break in the fencing and the wildlife guard. Though not all animals that
investigated the wildlife guard may have intended to cross the wildlife guard, an analysis that only included
these types of observations was less likely to overestimate the barrier effect of the wildlife guards than an
analysis that was solely based on the proximity of an animal to a wildlife guard.
Though we were not able to implement a formal before– after control impact study design, we did use data
obtained from sand tracking beds immediately adjacent to the high- way in earlier years (Jun–Oct of 2003–
2005) along the same road section before the highway was reconstructed and the mitigation measures were
implemented (i.e., there was no wildlife fencing and the animals could approach the road anywhere). The
1-km-long road section in Ravalli Curves had 8 sand tracking beds installed at random locations
immediately adjacent and parallel to the road (Hardy et al. 2007). Each sand tracking bed was 100 m long
and 2 m wide. We calculated and compared ‘‘turn around’’ and successful crossing rates on these sand
tracking beds with the crossing rates at the wildlife guards. For these analyses, we only included data for
deer and black bear, because observers were not able to reliably detect tracks of coyotes in the exposed
track bed substrate.
For the sand bed tracking study, observers recorded animal tracks as crossing the sand tracking beds,
moving parallel to the road, or as being simply present on the sand tracking beds. We assumed that the
animals crossed the road when their tracks were recorded as crossing and that the animals turned around
and did not cross the road when their tracks were recorded as parallel or present. We compared the preconstruction tracking bed data with the more recent wildlife guard data based on the number of individual
animals (i.e., not groups) that approached the wildlife guards within 2 m. We counted individuals because
we could not measure the number of groups for the sand bed tracking data; we could only measure the
number of individuals. The tracking data correspond to a 2-m-wide zone adjacent to U.S. Highway 93,
which was consistent with the wildlife guard data that correspond to animals that came within 2 m of the
2
test for a potential difference in the number of animals that turned around at the roadside tracking beds
(pre-construction sample) versus the number of animals that turned around at the wildlife guard.
Changes, especially increases, in traffic volume may act as a barrier to wildlife (Forman and Alexander
1998). We accounted for this potential confounding factor by analyzing traffic data from a traffic counter
located within the 1-km study area (MDT 2008, 2009, 2010). We used 2 years of traffic data (Aug 2008–
Jul 2010) that most closely corre- sponded to the 2 years that we monitored the wildlife guards
post-construction. Traffic data corresponding to the years we monitored road crossings pre-construction
were not avail- able, so we compared post-construction traffic data with the available traffic data that were
closest in time (1999–2000) to our pre-construction monitoring.
We predicted that the wildlife guards would be a substan- tial barrier to deer but would not be a substantial
barrier to black bear and coyotes, though it would be desirable to also reduce the number of black bear and
coyotes entering the fenced right-of-way. Other studies have shown that similar wildlife guards may be
effective barriers to deer movement (Peterson et al. 2003, Sebesta et al. 2003). Due to the design of the
grate and given the size of a black bear’s foot, we predicted bears could easily walk over the wildlife
guards. Similarly, due to their foot morphology and their agility, we predicted that coyotes would not have
great difficulty walk- ing over the wildlife guards.
We compared the number of wildlife movements across the northern-most wildlife guard toward U.S.
Highway 93 (E to W) with the number of wildlife movements from east to west through an adjacent
culvert. Our null hypothesis was that wildlife would use the wildlife guard and nearby crossing structure
equally. Our alternative hypothesis was that wildlife would use the wildlife guard less than the crossing
structure. To test these hypotheses, we performed a chi-square test.
RESULTS
We included species for which we had a sample size >10 in our analyses, with the exception of one
analysis done for black bears with a sample size of 9. For all 3 species (deer, black bear, and coyotes) and
all analyses, all groups of animals stayed together, either all crossing or not crossing as a group.
We counted 137 independent animal records for all species approaching the 2 wildlife guards within 2 m
and 65 (47.4%) of these resulted in a crossing of a wildlife guard. The species observed included deer,
black bear, domestic cat, bobcat (Lynx rufus), mountain lion (Puma concolor), fox (Vulpes vulpes), coyote,
wolf (Canis lupus), domestic horses (both alone and with a human rider), raccoon (Procyon lotor), and
skunk (Mephitis mephitis). Most deer were deterred from crossing the wildlife guards and some black bear
and coyotes were also discouraged from crossing (Table 2).
We observed 104 independent animal groups (all species) approaching a wildlife guard and displaying
behavior indi- cating intent to cross. This is, by definition, a subset of the animals that approached the
wildlife guards within 2 m. Most (62.5%) of this subset of approaches resulted in a crossing of a wildlife
guard. The species observed included deer, black bear, domestic cat, bobcat, mountain lion, fox, coyote,
wolf, domestic horse (with human rider), raccoon, and skunk. Most deer did not cross the wildlife guards
and some black bear and coyotes were also deterred from crossing (Table 3).
During pre-construction monitoring, observers recorded 107 deer approaches in sand beds along the
roadside and 43.9% resulted in a road crossing. This is a larger percentage than the 5.9% of the approaches
(n 1⁄4 85) to the wildlife guard that resulted in a crossing post-construction
�
405
Table 2. Number and percentage of wildlife groups that approached (within 2 m) and subsequently crossed, or did not cross, wildlife
guards south of Ravalli, Montana, USA, during the 2-year study period (mid-Jul 2008–mid-Jul 2010).
Species
Approached Crossed % Crossed
Did not cross
30 3 34 5 12
% Effective
93.8 60.0 89.5 45.5 54.5
Mule deer
White-tailed deer
a
Black bear
Coyote 22
32 2 6.3 5 2 40.0 38 4 10.5 11 6 54.5 10 45.5
a
1
2
roadside pre- construction, 86.7% of which resulted in a road crossing. Post-construction, about the same
1
2
1⁄4 2.345, P 1⁄4 0.126).
Our null hypothesis that wildlife would use the wildlife guard and nearby crossing structure equally was
rejected in favor of the alternative that wildlife would use the wildlife guard less than the nearby crossing
structure. We observed 289 independent animal movements crossing either through the crossing structure
(88.2%) or over the wildlife guard (12.8%). We observed deer, domestic cat, bobcat, fox, coyote, and
raccoon both crossing the wildlife guard and going through the culvert. The species we observed only
going through the culvert included black bear, domestic dog (both with and without a human present),
beaver (Castor canadensis), river otter (Lontra canadensis), mountain cotton- tail (Sylvilagus nuttallii), and
duck and goose (Anatidae). Mountain lion, wolf, and domestic horse (with human rider) crossed only via
the wildlife guard.
We observed 46 independent deer crossings, 93.5% of which were through the crossing structure, whereas
1
2
black bear crossings (n 1⁄4 48) were through the crossing structure. Similarly, most (94.7%) of the
independent coyote crossings (n 1⁄4 57) occurred through the crossing structure, as com- pared with the
1
2
DISCUSSION
Though the wildlife guards we monitored were not an absolute barrier to deer, they were �85% effective in
securing
the fenced right-of-way. The wildlife guards were less effec- tive for black bear and coyotes (33–46% and
33–55%, re- spectively), but these species were apparently still somewhat deflected from crossing the
wildlife guards, especially where there was a crossing structure adjacent to the wildlife guard. Other studies
that investigated the effectiveness of wildlife guards had variable results. Some of the variation in results is
due to different designs for the wildlife guards, though the species or subspecies (body size), the animal’s
motivation to cross, and potential disturbance (e.g., presence or absence of an adjacent highway with
traffic) are also likely to be impor- tant parameters that influenced wildlife guard effectiveness.
The wildlife guard design studied by Reed et al. (1974) consisted of flat mill steel rails (1.3 cm wide at the
top) positioned parallel to each other every 10.2 cm. This design was a poor barrier to mule deer because 16
out of 18 (89%) wild-caught mule deer crossed the wildlife guard after they were released from their cage.
The design tested in our study consisted of steel grate and proved to be a much more effective barrier to
deer than the design tested by Reed et al. (1974). However, the deer that were released in Reed’s study
were likely very motivated to cross the wildlife guard and leave the area, whereas our observations relate to
a real road setting with comparatively unstressed deer.
Peterson et al. (2003) tested 3 different wildlife guard designs on Florida Key deer (O. v. clavium) in a
situation where wild deer had to cross the wildlife guard to get to a food source. Each design consisted of
steel grate with varying mesh sizes or orientation of the mesh. The 2 designs that had similar mesh sizes to
the grates used in our study were 75% effective for Florida Key deer. The third design included a diagonal
strip inside the meshes and had the highest barrier effect to Florida Key deer (99.5%). The results reported
by Peterson et al. (2003) were similar to our findings. However,
Table3.
Numberandpercentageofwildlifegroupsthatshowedintenttocrossbehaviorallyandsubsequentlycrossed,ordidnotcross,2wildlifeguardssout
hof Ravalli, Montana, USA, during the 2-year study period (mid-Jul 2008–mid-Jul 2010).
Species
Approached Crossed % Crossed
Did not cross
Mule deer
White-tailed deer
a
Black bear
Coyote 15
21 2 9.5 4 2 50.0 26 4 15.4 9 6 66.7 10 66.7
19 2 22 3 5
% Effective
90.5 50.0 84.6 33.3 33.3
�
a
we hypothesize that a design where a diagonal strip is added to the 8-cm � 10-cm meshes may be less of a
barrier to the much larger white-tailed deer in our study area than to the Florida Key deer.
There were some differences in our monitoring methods pre-construction versus post-construction (i.e.,
sand beds vs. cameras) and some potential confounding factors (e.g., traffic volume, the road was wider
post-construction). Traffic vol- ume decreased slightly after construction, and we observed that deer
crossed the wildlife guards far less than they crossed U.S. Highway 93 when no wildlife guards were
present during pre-construction. Though not as definitive as results from a before–after control impact
design, our results indicate the wildlife guards were functioning as de- sired in securing the fenced right-ofway and the wildlife guards were more of a barrier to deer than just the potential barrier effects of the road
and traffic itself. For black bears, this comparison did not result in a statistical difference. Although our
sample size was small, the data support our prediction that the wildlife guards were not a substantial barrier
to black bears.
Most deer, black bear, and coyotes crossed through a nearby crossing structure instead of using the wildlife
guard but when a crossing did occur over either of the wildlife guards, most animals simply walked across
the wildlife guard surfaces. However, deer both walked and jumped across, landing on the wildlife guard at
least once. Despite jumping on and walking across the grate, no deer or other animals appeared to be
injured while crossing or attempting to cross. We did not observe any animals completely jumping across
in one leap, which indicated that the wildlife guards are of sufficient length. Some animals did use the
concrete ledges at the edges of the wildlife guards to cross. However, when presented with 2 options, most
animals crossed the road using the crossing structure rather than the nearby wildlife guard, which indicated
that the crossing structure was po- tentially perceived as a preferred pathway by many of the animals.
MANAGEMENT IMPLICATIONS
Gaps in a fence, even when mitigated with wildlife guards, may reduce the barrier effect of the fence.
However, in developed areas where gaps for access roads are a necessity, we found wildlife guards to be an
effective means of mitiga- tion to reduce wildlife access to the fenced right-of-way, particularly where deer
are the main species of concern. If black bears, coyotes, and other wildlife that have foot mor- phology and
the agility to cross a wildlife guard are the target species for mitigating, different or additional measures
should be considered. One such additional measure is instal- ling a wildlife crossing structure in the
immediate vicinity of wildlife guards. Our data indicate this may increase the effectiveness of the wildlife
guard as a barrier.
Another measure is electric matting. To our knowledge it has not been tested specifically for bears or
coyotes, but it has been tested and found effective for deer. However, it is likely more costly over time than
wildlife guards. The concrete ledges in the wildlife guard design we tested slightly
increased the number of wildlife that were able to cross the wildlife guard. However, this may be mitigated
by adding chain-link fencing or mesh at a diagonal to cover the concrete ledge or bringing the fence over
the concrete ledge in its entirety. Another idea, which is currently being tested, is the placement of rubber
‘‘bumpers’’ that inhibit access to the concrete ledge (P. Basting, Montana Department of Transportation,
personal communication). Perhaps these or similar measures will further increase the effectiveness of the
wildlife guards.
ACKNOWLEDGMENTS
The authors would like to thank both of our funding sources: Research and Innovative Technologies
Administration, U.S. Department of Transportation, and S&K Electronics. We appreciate the cooperation
of the Confederated Salish and Kootenai Tribes and their willingness to let us conduct research on the
reservation. In particular, we would like to thank D. Becker and W. Camel-Means. We would also like to
thank the Montana Department of Transportation, espe- cially P. Basting, T. Callahan, and C. Smith. We
would also like to thank S. Creel for his help and input. In addition, we greatly appreciate the help of J.
Purdum and C. R. Allen.
LITERATURE CITED
Bashore, T. L., and E. D. Bellis. 1982. Deer on Pennsylvania airfields: problems and means of control. Wildlife Society Bulletin
10:386–388. Beckmann, J. P., A. P. Clevenger, M. P. Huijser, J. A. Hilty, editors. 2010.
Safe passages. Highways, wildlife, and habitat connectivity. Island Press,
Washington, D.C., USA.
Belant, J. L., T. W. Seamans, and C. P. Dwyer. 1998. Cattle guards reduce
white-tailed deer crossings through fence openings. International Journal
of Pest Management 44:247–249.
Bissonette, J. A., and M. Hammer. 2000. Effectiveness of earthen return
ramps in reducing big game highway mortality in Utah. Utah Cooperative
Fish and Wildlife Research Unit, Report Series 2000:1–29.
Clevenger, A. P., B. Chruszcz, and K. E. Gunson. 2001. Highway mitiga- tion fencing reduces wildlife–vehicle collisions. Wildlife
Society Bulletin
29:646–653.
Conover, M. R., W. C. Pitt, K. K. Kessler, T. J. DuBow, and W. A.
Sanborn. 1995. Review of human injuries, illnesses, and economic losses caused by wildlife in the United States. Wildlife Society
Bulletin 23:407– 414.
Dodd, N. L., J. W. Gagnon, S. Boe, A. Manzo, and R. E. Schweinsburg. 2007. Evaluation of measures to minimize wildlife–vehicle
collisions and maintain permeability across highways: Arizona Route 260. Arizona Department of Transportation, Final Report 540,
FHWA-AZ-07-540, Phoenix, USA.
Falk, N. W., H. B. Graves, and E. D. Bellis. 1978. Right-of-way fences as deer deterrents. Journal of Wildlife Management 42:646–
650.
Farrell, J. E., L. R. Irby, and P. T. McGowen. 2002. Strategies for ungulate– vehicle collision mitigation. Intermountain Journal of
Sciences 8:1–18. Federal Highway Administration. 2007. Highway statistics 2007. Lane
miles by functional system. http://www.fhwa.dot.gov/policyinformation/statistics/hm60_summary.cfm. Accessed 14 Oct 2010.
Forman, R. T. T., and L. E. Alexander. 1998. Roads and their major
ecological effects. Annual Review of Ecological Systems 29:207–231. Hardy, A., J. Fuller, M. P. Huijser, A. Kociolek, and M. Evans.
2007. Evaluation of wildlife crossing structures and fencing on US Hwy 93 Evaro to Polson, phase I: preconstruction data collection
and finalization of evaluation plan. Montana Department of Transportation, final report, FHWA/MT-06-008/1744-2, Helena, USA.
http://www.mdt.mt.gov/
research/projects/env/wildlife_crossing.shtml. Accessed: 3 Jan 2013. Huijser, M. P., J. W. Duffield, A. P. Clevenger, R. J. Ament, and
P. T. McGowen. 2009. Cost–benefit analyses of mitigation measures aimed at
�
407
reducing collisions with large ungulates in the United States and Canada: a decision support tool. Ecology and Society 14: http://www.
ecologyandsociety.org/vol14/iss2/art15/. Accessed 7 Oct 2009.
Huijser, M. P., P. McGowen, J. Fuller, A. Hardy, A. Kociolek, A. P. Clevenger, D. Smith, and R. Ament. 2008. Wildlife–vehicle
collision reduction study. Report to congress. U.S. Department of Transportation, Federal, Highway Administration, Washington,
D.C., USA. http:// www.tfhrc.gov/safety/pubs/08034/index.htm. Accessed 3 Jan 2013.
Montana Department of Transportation [MDT]. 2008. Montana’s automatic traffic recorders.
http://www.mdt.mt.gov/publications/docs/ datastats/atr/atrbook08.pdf. Accessed 11 Apr 2012.
Montana Department of Transportation [MDT]. 2009. Montana’s automatic traffic recorders.
http://www.mdt.mt.gov/publications/docs/ datastats/atr/atrbook09.pdf. Accessed 11 Apr 2012.
Montana Department of Transportation [MDT]. 2010. Montana’s automatic traffic recorders.
http://www.mdt.mt.gov/publications/docs/ datastats/atr/atrbook10.pdf. Accessed 11 Apr 2012.
Peterson, M. N., R. R. Lopez, N. J. Silvy, C. B. Owen, P. A. Frank, and A. W. Braden. 2003. Evaluation of deer-exclusion grates in
urban areas. Wildlife Society Bulletin 31:1198–1204.
Reed, D. F., T. D. I. Beck, and T. N. Woodard. 1982. Methods of reducing deer–vehicle accidents: benefit–cost analysis. Wildlife
Society Bulletin 10:349–354.
Reed, D. F., T. M. Pojar, and T. N. Woodard. 1974. Mule deer responses to deer guards. Journal of Range Management 27:111–113.
Sebesta, J. D., S. W. Whisenant, R. R. Lopez, and N. J. Silvy. 2003. Development of a deer-guard prototype for Florida Key deer.
Proceedings of the Annual Conference, Southeast Fish and Wildlife Agencies 57:337–347.
VerCauteren, K. C., N. W. Seward, M. J. Lavelle, J. W. Fischer, and G. E. Phillips. 2009. Deer guards and bump gates for excluding
white-tailed deer from fenced resources. Human–Wildlife Conflicts 3:145–153.
Associate Editor: Nielsen.
408
�
Copyright of Wildlife Society Bulletin is the property of Wiley-Blackwell and its content
may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or
email articles for individual use.