Contributed Paper

Demographic Connectivity for Ursid Populations at
Wildlife Crossing Structures in Banff National Park
MICHAEL A. SAWAYA,∗ ANTHONY P. CLEVENGER,† AND STEVEN T. KALINOWSKI‡
∗

Western Transportation Institute and Department of Ecology, Montana State University, Bozeman, MT 59717, U.S.A.
email sawaya.mike@gmail.com
†Western Transportation Institute, Montana State University, Bozeman, MT 59717, U.S.A.
‡Department of Ecology, Montana State University, Bozeman, MT 59717, U.S.A.

Abstract: Wildlife crossing structures are one solution to mitigating the fragmentation of wildlife populations
caused by roads, but their effectiveness in providing connectivity has only been superficially evaluated.
Hundreds of grizzly (Ursus arctos) and black bear (Ursus americanus) passages through under and overpasses
have been recorded in Banff National Park, Alberta, Canada. However, the ability of crossing structures to
allow individual and population-level movements across road networks remains unknown. In April 2006,
we initiated a 3-year investigation into whether crossing structures provide demographic connectivity for
grizzly and black bears in Banff National Park. We collected hair with multiple noninvasive methods to
obtain genetic samples from grizzly and black bears around the Bow Valley. Our objectives were to determine
the number of male and female grizzly and black bears that use crossing structures; examine spatial and
temporal patterns of crossings; and estimate the proportions of grizzly and black bear populations in the
Bow Valley that use crossing structures. Fifteen grizzly (7 female, 8 male) and 17 black bears (8 female, 9
male) used wildlife crossing structures. The number of individuals detected at wildlife crossing structures was
highly correlated with the number of passages in space and time. Grizzly bears used open crossing structures
(e.g., overpasses) more often than constricted crossings (e.g., culverts). Peak use of crossing structures for both
bear species occurred in July, when high rates of foraging activity coincide with mating season. We compared
the number of bears that used crossings with estimates of population abundance from a related study and
determined that substantial percentages of grizzly (15.0% in 2006, 19.8% in 2008) and black bear (17.6% in
2006, 11.0% in 2008) populations used crossing structures. On the basis of our results, we concluded wildlife
crossing structures provide demographic connectivity for bear populations in Banff National Park.
Keywords: connectivity, fragmentation, noninvasive genetic sampling, populations, roads, Ursus americanus,
Ursus arctos, wildlife crossing structures
´
Conectividad Demogr´afica para Poblaciones de Ursidos
en Estructuras para Cruce de Vida Silvestre en el Parque
Nacional Banff

Resumen: Las estructuras para el cruce de vida silvestre son una soluci´on para mitigar la fragmentaci´on de
poblaciones de vida silvestre provocada por los caminos, pero su efectividad para proporcionar conectividad
ha sido evaluada solo superficialmente. Se han registrado cientos de pasadas de osos grizzli (Ursus arctos)
y osos negros (U. americanus) en los cruces del Parque Nacional Banff, Alberta, Canad´
a. Sin embargo, se
desconoce la capacidad de las estructuras de cruce para permitir movimientos individuales y a nivel poblaci´
on.
En abril 2006, iniciamos una investigaci´
on de 3 a˜
nos para conocer si las estructuras de cruce proporcionan
conectividad demogr´
afica para osos grizzli y negros en el Parque Nacional Banff. Recolectamos pelo mediante
m´
ultiples m´etodos no invasivos para obtener muestras gen´eticas de osos grizzli y negros alrededor del Valle
Bow. Nuestros objetivos fueron determinar los patrones espaciales y temporales de los cruces; y estimar las
proporciones de las poblaciones de osos grizzli y negros que utilizan estructuras de cruce. Quince osos grizzli
(7 hembras, 8 machos) y 17 osos negros (8 hembras, 9 machos) utilizaron estructuras de cruce. El n´
umero de
individuos detectados en las estructuras de cruce se correlacion´
o estrechamente con el n´
umero de pasadas en
espacio y tiempo. Los osos grizzli utilizaron estructuras de cruce abiertas (e.g., puentes) m´
as frecuentemente

Paper submitted January 7, 2012; revised manuscript accepted August 17, 2012.

721
Conservation Biology, Volume 27, No. 4, 721–730
"
C 2013 Society for Conservation Biology
DOI: 10.1111/cobi.12075

 722

Wildlife Crossing Structures

que los cruces compactos (e.g., alcantarillas). El pico del uso de estructuras de cruce para ambas especies de osos
ocurri´
o en julio, cuando las altas tasas de actividad forrajera coinciden con la temporada de apareamiento.
Comparamos el n´
umero de osos que utilizaron cruces con estimaciones de la abundancia poblacional de un
estudio relacionado y determinamos que porcentajes sustanciales de las poblaciones de osos grizzli (15.0%
en 2006, 19.8% en 2008) y de osos negros (17.6% en 2006, 11.0% en 2008) utilizaron las estructuras de
cruce. Con base en nuestros resultados, concluimos que las estructuras de cruce proporcionan conectividad
demogr´
afica a poblaciones de osos en el Parque Nacional Banff.

Palabras Clave: caminos, conectividad, estructuras para el cruce de vida silvestre, fragmentaci´on, muestreo
gen´etico no invasivo, poblaciones, Ursus americanus, Ursus arctos

Introduction
Road networks are one of the most ubiquitous anthropogenic features on the planet, and they pose a major
threat to Earth’s biodiversity (Forman et al. 2003). The
presence of roads is highly correlated with changes in
species composition, population sizes, and ecological
processes (Trombulak & Frissell 2000). Roadways, particularly ones with higher traffic volumes, can negatively
affect wildlife populations through direct mortality and
habitat fragmentation (Fahrig & Rytwinski 2009). Traffic
on roads is a substantial source of mortality for many
vertebrate populations, but road-caused mortality rates
rarely limit population abundance. Barrier effects caused
by road avoidance are thought to be a much bigger
ecological problem (Forman & Alexander 1998). Roads
constrain wildlife movements and thus decrease landscape connectivity and disrupt ecosystem functions such
as individual survival and reproduction, population persistence and evolution, predator–prey dynamics, and resource competition. Provision of wildlife crossing structures combined with exclusion fencing is one solution
to keeping animals off roadways while maintaining or
restoring animal movements, but the ability of wildlife
crossings to provide connectivity for populations fragmented by roads has yet to be properly evaluated (Corlatti
et al. 2009; Kaplan 2009).
Patches of habitat that are connected will have greater
species richness, diversity, and persistence than isolated
patches (MacArthur & Wilson 1967). Corridors are designed to link patches of fragmented habitat and thus to
stabilize population dynamics and community processes.
Wildlife crossing structures are “in essence site-specific
movement corridors” that are “designed to link critical
habitats and provide safe movement of animals across
busy roads” (Clevenger & Wierzchowski 2006). Potential
demographic benefits of mitigating roads with crossing
structures include increased survival from lower road
mortality, increased survival and fecundity from greater
access to patchily distributed resources (i.e., food and
shelter), and increased recruitment from emigration. Ecological connectivity will become increasingly important
as climate changes (Krosby et al. 2010). Wildlife crossings
may help mitigate some effects of climate change by al-

Conservation Biology
Volume 27, No. 4, 2013

lowing animals to change movements (i.e., latitude, elevation) in response to rising temperatures and by allowing
the flow of genes so that populations can adapt to and
evolve with changing environmental conditions. Despite
these potential benefits, conservation biologists have
hotly debated the conservation value of corridors, including engineered wildlife crossing structures, for over 20
years (e.g., Noss 1987; Simberloff & Cox 1987; Beier &
Noss 1998). The debate has understandably focused on
whether wildlife crossings and other landscape corridors
designed to increase population viability are worth the
relatively high costs (e.g., Hobbs 1992; Simberloff et al.
1992; Dixon et al. 2006).
The ability of wildlife crossing structures to connect
populations remains surprisingly unknown considering
there is an aggressive global campaign currently underway to incorporate crossing structures into construction
projects (Forman 2000; Ament et al. 2008; Mata et al.
2008). Exclusion fencing can reduce wildlife–vehicle collisions for some species (Clevenger et al. 2001; Jaeger
& Fahrig 2004), and a variety of species use wildlife
crossings to traverse busy highways (Clevenger & Waltho
2000; Clevenger & Waltho 2005). Nonetheless, there
have been few studies on the effectiveness of crossing
structures that go beyond documentation of the number
of species that use them and their frequency of use (e.g.,
Cain et al. 2003; Ng et al. 2004; Gagnon et al. 2011).
One of the world’s most developed and well-known
systems of wildlife crossings is located in Banff National
Park (BNP) in the Central Rocky Mountains of Alberta,
Canada. With over 4 million visitors per year, BNP is
one of the world’s most visited national parks and this
high level of human visitation acts as a major stressor on
the ecosystem (Ford et al. 2010). Some of the seminal
publications on road ecology have come from studying
the 2 overpasses (>US $2 million/structure) and 23 underpasses built in the 1980s and 1990s to reduce wildlife–
vehicle collisions and restore or maintain wildlife movement across the 4-lane section of the Trans-Canada Highway (TCH) (Clevenger & Waltho 2000; Clevenger et al.
2001; Clevenger & Waltho 2005). The TCH is Canada’s
busiest highway (17,970 vehicles/day on average in BNP)
and the nation’s primary east–west transportation route
(Ford et al. 2010). Given the extent and duration of

 Sawaya et al.

road-mitigation efforts in BNP along with its protected
status and biodiversity, BNP is viewed the world over as
the quintessential place to study the effectiveness of road
mitigation measures.
Recent evidence indicates that healthy populations of
apex consumers (e.g., carnivores) are critical to maintaining top-down ecological processes, but many carnivore species are declining rapidly worldwide (Estes et al.
2011). Wide-ranging, large-bodied carnivores that inhabit
BNP such as grizzly (Ursus arctos) and black bears (Ursus
americanus) are susceptible to road-caused population
fragmentation due to their low densities and reproductive rates and large home ranges (Brody & Pelton 1989;
Proctor et al. 2005; Rytwinski & Fahrig 2011). Wildlife
crossing structures in BNP and the bears that use them
represent the ideal study system and model study species
to address huge gaps in our understanding of how, when,
and why road mitigation measures provide connectivity
for individuals and populations.
Until recently, remote cameras and track pads have
been the most frequently used methods to evaluate the
effectiveness of wildlife crossing structures (Long et al.
2008; Ford et al. 2009), but these methods cannot reliably
distinguish individuals or genders, so the research questions they can address are limited (Clevenger & Sawaya
2010). For instance, hundreds of passages of grizzly and
black bears through wildlife crossing structures (hereafter passages) have been recorded with track pads at
wildlife crossing structures in BNP. However, due to
methodological limitations, the ability of crossing structures to serve as movement corridors for multiple individuals of different sexes remains completely unknown.
Recent interest in road ecology has focused on the use of
molecular genetics to investigate transportation effects
on wildlife (Balkenhol & Waits 2009). Samples collected
via noninvasive genetic sampling can identify male and
female individuals and may provide information on abundance, vital rates, and genetic interchange, but managers
have been reluctant to embrace genetic-monitoring methods because they are relatively new (Schwartz et al.
2006). Genetic data have been used frequently to assess genetic connectivity, but they may also be used to
assess demographic connectivity, the degree to which
population growth and vital rates are affected by individual movements between populations (Lowe & Allendorf
2010). A new noninvasive genetic sampling method allows assessment of the ability of wildlife crossings to
connect ursid populations (Clevenger & Sawaya 2010).
In April 2006, we initiated a 3-year evaluation of the
effectiveness of wildlife crossing structures to provide
demographic connectivity for grizzly and black bear populations in the Bow Valley of BNP. We used the hair
collection method developed by Clevenger and Sawaya
(2010) to sample crossing structures. In a related study,
Sawaya et al. (2012) used a combination of hair traps
and rub tree surveys to collect genetic samples from
grizzly and black bears in the Bow Valley. Here, our

723

main objectives were to determine the minimum number of individual male and female grizzly and black bears
that use wildlife crossing structures to traverse the TCH,
examine the spatial and temporal patterns of individual
bear movements at crossing structures, and estimate the
proportions of the bear populations in the Bow Valley
that use crossing structures to traverse the TCH.

Methods
Study Area
Our study area in the Bow Valley of BNP was 2246 km2
and located approximately 120 km west of Calgary, Alberta, east of the Continental Divide in the central Rocky
Mountains (Supporting Information). The lower Bow Valley is a human-dominated landscape with the TCH, the
Banff Townsite (8000 residents), a golf course, 3 ski areas,
a railway, and a secondary highway. Between 1982 and
1997, a 45-km stretch of TCH, extending west from the
eastern park boundary, was widened from 2 to 4 lanes
for safety reasons (McGuire & Morrall 2000). Twentyfive wildlife crossing structures, 2 overpasses and 23
underpasses (e.g., box culverts, creek, and open-span
bridges), and 2.4-m high fencing were constructed to
facilitate wildlife movement and reduce wildlife–vehicle
collisions, respectively, along the 4-lane section of the
TCH (Ford et al. 2010). Detailed ecological descriptions
of the study area are in Holroyd and Van Tighem (1983)
and Holland and Coen (1983).
Hair Sampling
We conducted hair sampling concurrently with track pad
monitoring at 20 of 25 wildlife crossing structures in the
Bow Valley from 15 May to 18 October 2006, 22 April
to 29 October 2007, and 22 April to 18 October 2008
(Supporting Information). Five wildlife crossings were
not included because they receive high levels of human
use and low levels of bear use (Clevenger & Waltho
2005). To collect hair samples we stretched 2 lengths
of barbed wire perpendicular to the line of movement of
passing bears (Clevenger & Sawaya 2010). We collected
hair samples in conjunction with track pad monitoring.
Track pads consisted of strips of sandy loam 1.5- to 2-m
wide perpendicular to the line of bear movement and
spanning the length of the wildlife crossing (Clevenger
& Waltho 2000; Clevenger & Waltho 2005). We checked
track pads every 2 days and recorded species, direction
of travel, and number in group (Ford et al. 2009). We
based species identification on diagnostic characteristics
of tracks (Halfpenny 2001). We calculated total number
of passages by summing discrete crossing events over the
3-year sampling period.
For hair sampling at the crossing structures, we secured 2 metal stakes at each end of a track pad, and
attached the 2 strands of wire at 30 cm and 70 cm,
Conservation Biology
Volume 27, No. 4, 2013

 Wildlife Crossing Structures

724

respectively. Page wire and brush were piled at the ends
of the wire to funnel animals through the hair-sampling
system. Illustrations and detailed descriptions of the hairsampling system and related pilot study are in Clevenger
and Sawaya (2010). We checked the barbed wire for hair
every 2 days in conjunction with track pad monitoring.
We collected hair samples from each barb separately
and collected discrete clumps of hair below the wire.
We placed samples in paper envelopes labeled with a
uniquely numbered barcode and sterilized every barb
from which we collected hair with an open flame to prevent DNA contamination. We stored hair samples at room
temperature in plastic bins containing silica desiccant.
Detailed descriptions of genetic analyses are available in
Supporting Information.
In a previous study, Sawaya et al. (2012) used 2 noninvasive genetic sampling methods, hair traps (Woods et al.
1999) and bear rubs (Kendall et al. 2009), to collect hair
from black and grizzly bear populations near the TCH and
to estimate grizzly and black bear population abundance
in the Bow Valley in 2006 and 2008. Detailed descriptions of hair collection, genetic analyses, and abundance
estimation methods and results used to estimate the proportions of the bear populations in the Bow Valley that
use crossing structures to traverse the TCH are in Sawaya
et al. (2012).
Demographic Connectivity
We considered a group of individuals of the same species
was demographically connected if they co-occurred in
space and time and had an opportunity to interact. We
based these criteria on the ecological paradigm definition of population cohesiveness proposed by Waples and
Gaggiotti (2006). We first examined individual patterns
of bear passages across wildlife crossings along the TCH.
We used multilocus genotypes to count the number of
unique male and female bears that used crossing structures each year and over the 3 years of sampling. Many
crossing events resulted in the collection of multiple hair
samples from the same individual, but were only counted
as single passages. Some individuals were detected at the
same wildlife crossing structure on multiple occasions,
but were counted only once per crossing structure. Some
individuals were repeatedly detected at the same type of
wildlife crossing structure, but were counted only once
per crossing structure type. Some individuals were detected more than once in a given month or during more
than 1 month in the same year or across years but were
counted only once per month or year. We calculated
mean and standard error for frequency of individual bear
passages by year and across years. We used graphs to visualize spatial (i.e., crossing structure) and temporal (i.e.,
month) patterns of individual bear use. We calculated
spatial and temporal correlation coefficients to determine
the relation between the number of passages recorded at
track pads and the number of individuals.
Conservation Biology
Volume 27, No. 4, 2013

We assessed population-level connectivity by comparing minimum counts of grizzly and black bears detected
at crossing structures with abundance estimates of grizzly
and black bears in the Bow Valley of BNP from Sawaya
et al. (2012). We estimated the minimum proportion of
each population that used crossing structures by dividing
the minimum number of male and female bears detected
at crossing structures in a year by the respective annual
superpopulation abundance estimates from Sawaya et al.
(2012) for each species and sex.

Results
Hair Sampling
We recorded 91 grizzly bear passages at track pads in
2006, 115 in 2007, and 180 in 2008 (Supporting Information). We recorded 129 black bear passages at track pads
in 2006, 98 in 2007, and 88 in 2008 (Supporting Information). We collected 348 hair samples from wildlife crossings in 2006, 416 in 2007, and 553 in 2008 (Supporting
Information). Many samples came from the same individual bears or nontarget species such as cougars (Puma
concolor) and wolves (Canis lupus). We collected hair
from all wildlife crossings (n = 20) monitored in all years,
but not all passages recorded at track pads yielded useable DNA or DNA from the target species. We identified
individuals from 37% of grizzly (n = 386) and 33% of
black bear passages (n = 315) recorded at track pads
between 2006 and 2008.
Demographic Connectivity
We identified 15 grizzly bears (8 males, 7 females) and
17 black bears (9 males, 8 females) that used the wildlife
crossing structures over the course of the study. We identified 11 grizzly and 11 black bears that used crossings in
2006, 12 grizzly and 8 black bears in 2007, and 10 grizzly
and 9 black bears in 2008 (Table 1). A high percentage
of individual grizzly bears detected at crossings were detected in >1 year (71% of females and 75% of males). A
lower percentage of black bears were detected in >1 year
(37% of females and 55% of males). Fifty-seven percent of
female and 37% of male grizzly bears were detected in all
3 years of sampling compared with only 25% of female
and 11% of male black bears. The average number of
passages per individual was higher for grizzly bears than
for black bears (Supporting Information). Variability in
individual passage frequency was high for both species.
Of the 247 total passages that produced useable DNA
samples, grizzly bear males had the highest number of
total passages (n = 97) and the highest mean number of
passages per individual (¯x [SD] = 12.1 [12.8]). One grizzly
bear male accounted for 34 of the 97 (35.1%) passages
by male grizzly bears for which we had useable DNA,
including the highest number we recorded for a single

 Sawaya et al.

725

Table 1. Number of individual bears detected in Bow Valley of Banff
National Park, Alberta, Canada, with 3 noninvasive genetic sampling
methods∗ (barbed wire hair sampling at wildlife crossings, hair traps,
and bears rubs) between April 2006 and October 2008.
Detection method
Wildlife crossings
grizzly bears
males
females
black bears
males
females
Hair traps
grizzly bears
males
females
black bears
males
females
Bear rubs
grizzly bears
males
females
black bears
males
females

2006

2007

2008

Total

11
6
5
11
7
4

12
6
6
8
4
4

10
5
5
9
4
5

15
8
7
17
9
8

31
16
15
40
16
24

N/A
N/A
N/A
N/A
N/A
N/A

39
12
17
57
19
38

42
20
22
76
30
46

40
25
15
2
1
1

46
30
16
11
9
2

43
24
19
6
2
4

73
44
29
16
12
4

∗ Some

individuals detected with more than one method or in more
than one year.

year, 17. One grizzly bear female accounted for 18 of 47
(38.3%) female grizzly bear passages. Single individuals
accounted for an even larger percentage of black bear
passages (24 of the 55 [43.6%] passages by 1 male and
21 of 48 [43.8%] female black bear passages by a single
female).
The number of grizzly and black bear passages was
spatially correlated with the number of unique individuals detected at wildlife crossing structures (Pearson’s
r = 0.89 for black bears and r = 0.95 for grizzly bears)
(Fig. 1). Grizzly bear passages and individuals were highly
concentrated at 2 overpasses and 1 open-span underpass
to the west of Banff townsite, whereas black bear passages were evenly distributed and more concentrated to
the east (Fig. 1). High proportions of grizzly bear males
and females were detected at the most open types of
wildlife crossings, overpasses and open-span underpasses
(Fig. 2a). Over 70% of female and 75% of male grizzly
bears detected at wildlife crossings over the study period
were detected at 1 of just 2 overpasses (Fig. 2a). The
proportion of black bears detected at each crossing type
was more evenly distributed (Fig. 2b).
The number of grizzly and black bears detected at
wildlife crossings during each month was highly correlated with the number of passages detected with track
pads (r = 0.99 for black bears and r = 0.98 for grizzly
bears) (Fig. 3). The number of grizzly bear individuals
identified that used wildlife crossings peaked in July, but
increased slightly in September (Fig. 3a). The peak in

Figure 1. Total number of grizzly and black bear (a)
passages through crossing structures and (b)
individuals detected at each wildlife crossing structure
between 2006 and 2008. Number of passages is the
sum of unique crossing events recorded at track pads
over 3 years. Number of individuals is the sum of
unique individuals detected at a given crossing
structure on the basis of DNA analysis of hair samples
over 3 years. Crossing structures: M, medium culvert
(n = 1); B, box culvert (n = 4); L, large culvert (n =
5); C, creek bridge (n = 3); S, open-span (n = 5); and
O, overpass (n = 2). Crossing structures are ordered
from west (left) to east (right).
number of black bear individuals and passages in July
was higher than the peak for grizzly bears (Fig. 3b).
From Sawaya et al. (2012), we obtained precise abundance estimates for grizzly and black bears during 2006
and 2008 in the Bow Valley from which we determined
that, in each year, >10% of the populations of both
species used wildlife crossings to traverse the TCH. We
estimated the minimum proportion of the grizzly bear
population that used crossings was 15.0% in 2006 and
19.8% in 2008 (Table 2). We estimated the minimum proportion of the black bear population that used crossings
was 17.6% in 2006 and 11.0% in 2008 (Table 3).

Discussion
To justify the continuation of expensive road mitigation
projects, transportation agencies and wildlife managers

Conservation Biology
Volume 27, No. 4, 2013

 Wildlife Crossing Structures

726

Figure 2. Proportion of total number of (a) grizzly
and (b) black bears detected at each of 6 types of
crossing structures monitored with noninvasive
genetic sampling between 2006 and 2008. Proportions
are based on the number of unique individuals
detected at a crossing type over 3 years. Number of
individuals is the sum of unique individuals detected
at a given crossing structure type; individual identity
was determined from DNA analysis of hair samples
collected at wildlife crossings over 3 years. Crossing
structures are ordered from more constricted (left) to
more open (right): medium culvert (n = 1), box
culvert (n = 4), large culvert (n = 5), creek bridge
(n = 3), open-span (n = 5), and overpass (n = 2).

Figure 3. Total number of individual (a) grizzly and
(b) black bears detected each month with noninvasive
genetic sampling at wildlife crossings in Banff
National Park, Alberta between 2006 and 2008.
Number of passages is the sum of unique crossing
events recorded at track pads over 3 years (species
identification on the basis of tracks). Number of
passages for which DNA samples were acquired (i.e.,
passages with DNA) is the sum of unique crossing
events that produced a hair sample with amplifiable
DNA. Number of individuals is the sum of unique
individuals detected at a given crossing structure;
individual identity was determined from DNA
analysis of hair samples collected at wildlife crossings
over 3 years.

Table 2. Total minimum number of grizzly bears and proportion of population detected using wildlife crossings to traverse the Trans-Canada
Highway in the Bow Valley of Banff National Park, Alberta, Canada, in 2006 and 2008.
Year and sex
2006
males
females
total
2008
males
females
total

Minimum
number

Abundance
a
estimate

ˆ)
SE ( N

No. detected
at crossings

Population detected
b
at crossings (%)

32
25
57

39.9
33.6
73.5

4.7
5.3
7.2

6
5
11

15.0
14.9
15.0

26
22
48

28.1
22.3
50.4

2.1
0.6
2.2

5
5
10

17.8
22.4
19.8

a Model-averaged abundance estimates from Sawaya et al. (2012).
b Estimated by dividing the number detected at the wildlife crossings

Conservation Biology
Volume 27, No. 4, 2013

by the abundance estimate.

 Sawaya et al.

727

Table 3. Total minimum number of black bears and proportion of population detected using wildlife crossings to traverse the Trans-Canada Highway
in the Bow Valley of Banff National Park, Alberta, Canada, in 2006 and 2008.
Year and sex
2006
males
females
total
2008
males
females
total

Minimum
number

Abundance
a
estimate

ˆ)
SE ( N

No. detected
at crossings

Population detected
b
at crossings (%)

19
24
43

30.2
32.3
62.6

7.3
4.9
9.0

7
4
11

23.2
12.4
17.6

22
41
63

28.3
53.5
81.8

4.1
5.7
7.2

4
5
9

14.1
9.3
11.0

a Model-averaged abundance estimates from Sawaya et al. (2012).
b Estimated by dividing the number detected at the wildlife crossings

need to know whether wildlife crossings provide connectivity for individuals and populations. Rarely has empirical data on individual use of wildlife crossings been
collected that provides population-level context to those
movements (Corlatti et al. 2009; Kaplan 2009). Ours is
the first in-depth examination of individual movements,
spatial and temporal patterns of use, and population-level
connectivity afforded by wildlife crossing structures. On
the basis of our data, we conclude that crossing structures
provide connectivity for ursid populations.
Although many of the hair samples we collected at
crossings could not be used in genetic analyses, we were
able to identify individuals from 37% of grizzly and 33%
of black bear passages; thus, we had a conservative minimum count of bears that used wildlife crossings to traverse the TCH.
An examination of spatial patterns of individual bear
use of wildlife crossing structures revealed that the type
of crossing structure is important in determining where
certain species are likely to cross. A much higher percentage of male and female grizzly bears used large, open
crossings than black bears, whereas black bears used
smaller, more constricted crossings more often than grizzly bears (Fig. 2). These findings are consistent with trackpad counts of Clevenger and Waltho (2005). Considering
the high use of individual grizzly bears of more open,
less constricted crossings, we recommend transportation
planners and engineers consider overpasses and openspan underpasses when constructing crossings for grizzly
bears. It appears that black bears are more adaptable and
use a wider variety of crossing types than grizzly bears
(Fig. 2), so mitigation targeted for black bears could involve a broader array of smaller crossing types.
Given the high degree of variability in the number
of passages per individual (Supporting Information), it
was surprising to find that the number of passages determined from track-pad counts was highly correlated with
the number of individuals for both bear species (Fig. 1).
This provides some independent validity to studies of
crossing structure effectiveness that make assumptions
about higher-order processes solely on the basis of track-

by the abundance estimate.

count data. Our results demonstrate that wildlife crossings provide spatial connectivity by allowing frequent
movements by a number of individuals in a variety of
locations and through a number of different types of
crossing structures.
We also found some interesting temporal patterns. The
peak in grizzly and black bear passages in July (Fig. 3) was
not surprising because bears would use crossings most in
midsummer because it corresponds with mating season,
dispersal, and the highest level of foraging activity in
the montane valley bottom habitat. We also predicted an
abrupt decline in bear use of the crossings as movements
decreased due to increased foraging efficiency after buffaloberry (Shepherdia canadensis) patches ripen in late
July (Hamer & Herrero 1987). The number of passages we
recorded per month at crossings also was highly correlated with the number of individuals detected per month
(Fig. 3).
The number of male and female black and grizzly bears
that used crossings was surprisingly consistent across
years (Table 1), especially when considering the upward
trend in grizzly bear passages and the recent downward
trend in black bear passages (Supporting Information).
These opposing trends are complicated by the apparent relation between the number of track-pad detections
and number of individuals (Figs. 1 & 3). The number
of grizzly and black bears that used passages was fairly
constant across years (Table 1). A very high percentage
of grizzly bears used wildlife crossings from year to year,
but we resampled far fewer black bears across years. The
upward trend in grizzly bear passages coupled with the
relatively stable number of individuals that used crossing
structures suggests the increase in grizzly bear use may
be explained by increasingly frequent use by a few individuals that disproportionately used crossing structures.
Previous researchers have proposed that there may be
a strong learning curve for some species to find and
use wildlife crossings, and the strong upward trend we
found (Supporting Information) shows the importance of
long-term monitoring of crossing structures (Clevenger &
Waltho 2005; Gagnon et al. 2011).

Conservation Biology
Volume 27, No. 4, 2013

 Wildlife Crossing Structures

728

We provided a population-level context to our wildlife
crossing data by obtaining precise estimates of abundance from a previous study, Sawaya et al. (2012), and
directly comparing estimates with the minimum number of individuals we detected with hair sampling. Results of Hastings’ (1993) examination of the complex
interactions between dispersal and population dynamics with logistic equations suggest that discrete patches
are demographically connected (i.e., coupled) if ≥10%
of the population migrates between patches (Waples &
Gaggiotti 2006; Lowe & Allendorf 2010). Relatively large
percentages of the populations of grizzly (15–20%) and
black bears (11–18%) in the Bow Valley used wildlife
crossings to traverse between the north and south sides
of the TCH. These minimum percentages are especially
conservative when considering that our counts of bears
that used crossings were absolute minimum numbers and
these numbers were divided by maximum population
sizes (i.e., superpopulation estimates). Also, many bears
are unlikely to have home ranges overlapping or bordering the TCH; therefore, most bears would have no need
to use the crossings to access food or mates. Regardless,
nearly equal proportions of males and females of both
species were detected at crossings in 2 separate years;
a particularly encouraging finding because Proctor et al.
(2005) found significant fragmentation effects and malebiased dispersal of grizzly bears across a major east–west
highway in Alberta and British Columbia, Canada.
Although the minimum numbers of individuals that
used crossings were relatively consistent from year to
year, the population estimates for grizzly bears decreased
and the estimates for black bears increased from 2006 to
2008. Changes in abundance explain much of the variability in minimum estimates of the proportions of individuals using crossings between years, but the opposing
trend shown in passages from track pads (Supporting
Information) reinforces the evidence that the number of
passages is driven by a few individuals, irrespective of
abundance. It is clear that it takes time for grizzly bears
to learn how and where to incorporate wildlife crossings
into home range movements. But as grizzly bears learn to
use crossing structures it appears they will use crossings
with more regularity. In situations where adult survival
is low (e.g., hunted populations), grizzly bears may not
survive long enough to learn to use wildlife crossings;
therefore, crossings may not be as effective for unprotected populations as they are for protected populations.
Ideally, wildlife crossings should increase the permeability of roads and, therefore, survival of animals that
live near them, but direct and indirect effects of road
mitigation are complicated and difficult to evaluate without premitigation data (Roedenbeck et al. 2007). Wildlife
crossings can positively affect bear survival by reducing
roadkill, increasing access to food and shelter, and allowing escape from predation by other bears. Repeated
visual observations of bears foraging in the small strip of

Conservation Biology
Volume 27, No. 4, 2013

riparian habitat between the highway and the Bow River
suggest wildlife crossings in BNP play a role in allowing
access to seasonal food resources that may be especially
important to one of the most fragmented, food-stressed,
and human-influenced populations of grizzly bears on
Earth (Gibeau et al. 2002; Chruszcz et al. 2003). Even
though just a few individuals accounted for the majority
of female grizzly and black bear crossings in our study,
their use of crossings could be crucial to maintaining bear
population viability because even slight changes in adult
female survival can strongly affect population growth
rates for grizzly and black bears in BNP (Hebblewhite
et al. 2003; Garshelis et al. 2005).
Indirect effects of road mitigation are rarely considered, but they could be as influential on population viability as direct effects. Stress can affect physiology, behavior, and demography of wildlife (Romero 2004). Animals
that use crossings may experience reduced stress, which
may lead to increased health (i.e., survival) and reproductive fitness. Survival may be negatively affected if wildlife
crossings allow movements into areas with high habitat
quality and high mortality (e.g., railways, urban developments). Wildlife crossings are no guaranteed solution
for landscape and species conservation and the potential merits of building each crossing should be weighed
against the costs and potential direct and indirect effects
of road mitigation.
On the basis of our results, we suggest wildlife crossings provide connectivity for grizzly and black bear populations in BNP. Further study is warranted to understand the ability of wildlife crossings to prevent genetic
isolation and enhance population viability. The oneto-ten-migrant-per-generation rule of thumb proposes a
threshold of genetic interchange that should minimize
short-term effects of inbreeding while allowing for local
adaptive differentiation (Mills & Allendorf 1996; Wang
2004). Although movement does not necessarily translate to the flow of genes (Riley et al. 2006; Strasburg
2006), wildlife crossings may help maintain genetic connectivity. In short, our results demonstrate the ability of
wildlife crossing structures to provide spatial and temporal connectivity for wildlife populations across a major
transcontinental transportation corridor.

Acknowledgments
Funding was provided by Parks Canada, Western Transportation Institute, Woodcock Foundation, H.P. Kendall
Foundation, Wilburforce Foundation, National Fish and
Wildlife Foundation, Alberta Conservation Association,
Calgary Foundation, and Mountain Equipment Cooperative. S. Kalinowski was supported by National Science Foundation grant #DEB 0717456. R. Ament secured
funds. S. Woodley and K. Rettie championed our research
within Parks Canada. We give special thanks to B. Fyten,

 Sawaya et al.

M. Gibeau, BNP wardens, and many field assistants for
their contributions in the office and field. D. Paetkau
produced reliable genetic data. J. Stetz, L.S. Mills, and S.
Creel provided valuable comments on early manuscript
drafts.

Supporting Information
Map of Bow Valley study area and sampling locations
in Banff National Park, Alberta (Appendix S1), summary of noninvasive genetic sampling effort and collection success (Appendix S2), description of methods of
genetic analyses and results (Appendix S3), number of
grizzly and black bear passages at wildlife crossing structures between 1997 and 2008 (Appendix S4), and descriptive statistics for frequency of passages at wildlife
crossing structures by individual grizzly and black bears
(Appendix S5). The authors are solely responsible for
the content and functionality of these materials. Queries
(other than absence of the material) should be directed
to the corresponding author.

Literature Cited
Ament, R., A. P. Clevenger, O. Yu, and A. Hardy. 2008. An assessment
of road impacts on wildlife populations in U.S. National Parks. Environmental Management 42:480–496. [Correction added after publication 1 July 2013: The publication title was updated for accuracy.]
Balkenhol, N., and L. P. Waits. 2009. Molecular road ecology: exploring
the potential of genetics for investigating transportation impacts on
wildlife. Molecular Ecology 18:4151–4164.
Beier, P., and R. F. Noss. 1998. Do habitat corridors provide connectivity? Conservation Biology 12:1241–1252.
Brody, A. J., and M. R. Pelton. 1989. Effects of roads on black bear
movements in western North Carolina. Wildlife Society Bulletin
17:5–10.
Cain, A. T., V. R. Tuovila, D. G. Hewitt, and M. E. Tewes. 2003. Effects
of a highway and mitigation projects on bobcats in southern Texas.
Biological Conservation 114:189–197.
Chruszcz, B., A. P. Clevenger, K. E. Gunson, and M. L. Gibeau. 2003.
Relationships among grizzly bears, highways, and habitat in the
Banff-Bow Valley, Alberta, Canada. Canadian Journal of Zoology
81:1378–1391.
Clevenger, A. P., B. Chruszcz, and K. E. Gunson. 2001. Highway mitigation fencing reduces wildlife-vehicle collisions. Wildlife Society
Bulletin 29:646–653.
Clevenger, A. P., and M. A. Sawaya. 2010. Piloting a non-invasive genetic
sampling method for evaluating population-level benefits of wildlife
crossing structures. Ecology and Society 15(1):7. Available from
http://www.ecologyandsociety.org/vol15/iss1/art7/.
Clevenger, A. P., and N. Waltho. 2000. Factors influencing the effectiveness of wildlife underpasses in Banff National Park, Alberta, Canada.
Conservation Biology 14:47–56.
Clevenger, A. P., and N. Waltho. 2005. Performance indices
to identify attributes of highway crossing structures facilitating movement of large mammals. Biological Conservation 121:
453–464.
Clevenger, A. P., and J. Wierzchowski. 2006. Maintaining and restoring
connectivity in landscapes fragmented by roads. Pages 502–535 in
K. R. Crooks and M. Sanjayan, editors. Connectivity conservation.
Cambridge University Press, New York.

729
Corlatti, L., K. Hacklander, and F. Frey-Roos. 2009. Ability of wildlife
overpasses to provide connectivity and prevent genetic isolation.
Conservation Biology 23:548–556.
Dixon, J., M. Oli, M. Wooten, T. Eason, H. McCown, J. Walter, and D.
Paetkau. 2006. Effectiveness of a regional corridor in connecting
two Florida black bear populations. Conservation Biology 20:155–
162.
Estes, J. A., et al. 2011. Trophic downgrading of planet Earth. Science
333:301–306.
Fahrig, L., and T. Rytwinski. 2009. Effects of roads on animal
abundance: an empirical review and synthesis. Ecology and
Society 14(1):21. Available from http://www.ecologyandsociety.
org/vol14/iss1/art21/.
Ford, A. T., A. P. Clevenger, and A. Bennett. 2009. Comparison of
methods of monitoring wildlife crossing structures on highways.
Journal of Wildlife Management 73:1213–1222.
Ford, A. T., A. P. Clevenger, and K. Rettie. 2010. The Banff Wildlife
Crossings Project: an international public-private partnership. Pages
157–172 in J. Beckmann, A. P. Clevenger, M. Huijser, and J. Hilty,
editors. Safe passages: highways, wildlife and habitat connectivity.
Island Press, Washington, D.C.
Forman, R. T. T. 2000. Estimate of the area affected ecologically by the
road system in the United States. Conservation Biology 14:31–35.
Forman, R. T. T., and L. E. Alexander. 1998. Roads and their major ecological effects. Annual Review of Ecology and Systematics 29:207–
232.
Forman, R. T. T., et al. 2003. Road ecology: science and solutions. Island
Press, Washington, D.C.
Gagnon, J. W., N. L. Dodd, K. S. Ogren, and R. E. Schweinsburg. 2011.
Factors associated with use of wildlife underpasses and overpasses
and importance of long-term monitoring. Journal of Wildlife Management 75:1477–1487.
Garshelis, D. L., M. L. Gibeau, and S. Herrero. 2005. Grizzly bear demographics in and around Banff National Park and Kananaski Country,
Alberta. Journal of Wildlife Management 69:277–297.
Gibeau, M. L., A. P. Clevenger, S. Herrero, and J. Wierzchowski. 2002.
Grizzly bear response to human development and activitites in
the Bow River Watershed, Alberta, Canada. Biological Conservation
103:227–236.
Halfpenny, J. C. 2001. Scats and tracks of the Rocky Mountains: a field
guide to the signs of seventy wildlife species. A falcon guide. Globe
Pequot Press, Guilford, Connecticut.
Hamer, D., and S. Herrero. 1987. Grizzly bear food and habitat in the
Front Ranges of Banff National Park, Alberta. Pages 199–213 in P.
Zager, editor. Bears: their biology and management. Vol. 7, a selection of papers from the seventh international conference on bear
research and management, Williamsburg, Virginia, USA, and Plitvice
Lakes, Yugoslavia, February and March 1986.
Hastings, A. 1993. Complex interactions between dispersal and dynamics: lessons from coupled logistic equations. Ecology 74:
1362–1372.
Hebblewhite, M., M. Percy, and R. Serrouya. 2003. Black bear (Ursus
americanus) survival and demography in the Bow Valley of Banff
National Park. Biological Conservation 112:415–425.
Hobbs, R. J. 1992. The role of corridors in conservation: Solution or
bandwagon? Trends in Ecology & Evolution 7:389–392.
Holland, W. D., and G. M. Coen. 1983. Ecological (biophysical) land
classification of Banff and Jasper National Parks. Volume I: summary.
Alberta Institute of Pedology, Edmonton, AB.
Holroyd, G. L., and K. J. Van Tighem. 1983. Ecological (biophysical)
land classification of Banff and Jasper national parks. Volume 3. The
wildlife inventory. Canadian Wildlife Service, Edmonton, Alberta.
Jaeger, J. A. G., and L. Fahrig. 2004. Effects of road fencing on population
persistence. Conservation Biology 18:1651–1657.
Kaplan, M. 2009. Uncertainty over animal crossings: Are
bridges over busy roads helping wildlife to breed? Nature News 114. Available from http://www.nature.com/
news/2009/090223/full/news.2009.114.html.
Conservation Biology
Volume 27, No. 4, 2013

 730
Kendall, K. C., J. B. Stetz, J. Boulanger, A. C. Macleod, D. Paetkau, and G.
C. White. 2009. Demography and genetic structure of a recovering
grizzly bear population. Journal of Wildlife Management 73:3–17.
Krosby, M., J. Tewksbury, N. Haddad, and J. Hoekstra. 2010. Ecological
connectivity for a changing climate. Conservation Biology 24:1686–
1689.
Long, R. A., P. Mackay, W. J. Zielinski, and J. C. Ray, editors. 2008. Noninvasive survey methods for carnivores. Island Press, Washington,
D.C.
Lowe, W. H., and F. W. Allendorf. 2010. What can genetics tell us about
connectivity? Molecular Ecology 19:3038–3051.
MacArthur, R. H., and E. O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, New Jersey.
Mata, C., I. Hervas, J. Herranz, F. Suarez, and J. E. Malo. 2008. Are
motorway wildlife passages worth building? Vertebrate use of roadcrossing structures on a Spanish motorway. Journal of Environmental Management 88:407–415.
McGuire, T. M., and J. F. Morrall. 2000. Strategic highway improvements to minimize environmental impacts within the Canadian
Rocky Mountain national parks. Canadian Journal of Civil Engineering 27:523–532.
Mills, L. S., and F. W. Allendorf. 1996. The one-migrant-per-generation
rule in conservation and management. Conservation Biology
10:1509–1518.
Ng, S. J., J. W. Doyle, R. M. Sauvajot, S. P. D. Riley, and T. J. Valone. 2004.
Use of undercrossings by wildlife in southern California. Biological
Conservation 115:499–507.
Noss, R. F. 1987. Corridors in real landscapes: a reply to Simberloff and
Cox. Conservation Biology 1:159–164.
Proctor, M. F., B. N. McLellan, C. Strobeck, and R. M. R. Barclay. 2005.
Genetic analysis reveals demographic fragmentation of grizzly bears
yielding vulnerably small populations. Proceedings of the Royal Society B 272:2409–2416.
Riley, S. P. D., J. P. Pollinger, R. M. Sauvajot, E. C. York, C. Bromley, T.
K. Fuller, and R. K. Wayne. 2006. A southern California freeway is
a physical and social barrier to gene flow in carnivores. Molecular
Ecology 15:1733–1741.

Conservation Biology
Volume 27, No. 4, 2013

Wildlife Crossing Structures

Roedenbeck, I. A., L. Fahrig, C. S. Findlay, J. E. Houlahan, J. A. G.
Jaeger, N. Klar, S. Kramer-Schadt, and E. A. Van der Grift. 2007.
The Rauischholzhausen agenda for road ecology. Ecology and
Society 12(1):11. Available from http://www.ecologyandsociety.
org/vol12/iss1/art11/.
Romero, L. M. 2004. Physiological stress in ecology: lessons from
biomedical research. Trends in Ecology & Evolution 19:249–255.
Rytwinski, T., and L. Fahrig. 2011. Reproductive rate and body size predict road impacts on mammal abundance. Ecological Applications
21:589–600.
Sawaya, M. A., J. B. Stetz, A. P. Clevenger, M. L. Gibeau, and S. T. Kalinowski. 2012. Estimating grizzly and black bear population abundance and trend in Banff National Park using noninvasive genetic
sampling. PLoS ONE 7:DOI:10.1371/journal.pone.0034777.
Schwartz, M. K., G. Luikart, and R. S. Waples. 2006. Genetic monitoring
as a promising tool for conservation and management. Trends in
Ecology & Evolution 22:25–33.
Simberloff, D., and J. Cox. 1987. Consequences and costs of conservation corridors. Conservation Biology 1:63–71.
Simberloff, D., J. A. Farr, J. Cox, and D. W. Mehlman. 1992. Movement
corridors: Conservation bargains or poor investments? Conservation
Biology 6:493–504.
Strasburg, J. L. 2006. Roads and genetic connectivity. Nature 440:875–
876.
Trombulak, S. C., and C. A. Frissell. 2000. Review of ecological effects of
roads on terrestrial and aquatic communities. Conservation Biology
14:18–30.
Wang, J. 2004. Application of the one-migrant-per-generation rule
to conservation and management. Conservation Biology 18:
332–343.
Waples, R. S., and O. Gaggiotti. 2006. What is a population? An empirical
evaluation of some genetic methods for identifying the number of
gene pools and their degree of connectivity. Molecular Ecology
15:1419–1439.
Woods, J. G., D. Paetkau, D. Lewis, B. L. McLellan, M. Proctor, and C.
Strobeck. 1999. Genetic tagging of free-ranging black and brown
bears. Wildlife Society Bulletin 27:616–627.