Assessment of pronghorn movements and strategies to promote highway permeability : US Highway 89

              Assessment of Pronghorn
Movements and Strategies to
Promote Highway
Permeability: US Highway 89
Final Report 619
April 2011
Arizona Department of Transportation
Research Center
The contents of this report reflect the views of the authors who are responsible for the
facts and the accuracy of the data presented herein. The contents do not necessarily
reflect the official views or policies of the Arizona Department of Transportation or the
Federal Highway Administration. This report does not constitute a standard,
specification, or regulation. Trade or manufacturers’ names which may appear herein are
cited only because they are considered essential to the objectives of the report. The U.S.
Government and the State of Arizona do not endorse products or manufacturers.
Research Center reports are available on the Arizona Department of Transportation’s
internet site.
Technical Report Documentation Page
1. Report No.
FHWA-AZ-10-619
2. Government Accession No. 3. Recipient's Catalog No.
5. Report Date
MAY 2010
4. Title and Subtitle
ASSESSMENT OF PRONGHORN MOVEMENTS AND STRATEGIES
TO PROMOTE HIGHWAY PERMEABILITY
U.S. Highway 89 6. Performing Organization Code
7. Authors
Norris L. Dodd, Jeffrey W. Gagnon, Scott Sprague, Susan Boe,
and Raymond E. Schweinsburg
8. Performing Organization Report No.
9. Performing Organization Name and Address 10. Work Unit No.
Arizona Game and Fish Department
Research Branch
5000 W. Carefree Highway
Phoenix, AZ 85068
11. Contract or Grant No.
ECS File No. JPA 07-004T
13.Type of Report & Period Covered
FINAL REPORT
November ‘06– December ‘09
12. Sponsoring Agency Name and Address
Arizona Department of Transportation
206 S. 17th Avenue
Phoenix, AZ 85007
ADOT Project Manager: Estomih Kombe
14. Sponsoring Agency Code
15. Supplementary Notes
Prepared in cooperation with the U.S. Department of Transportation, Federal Highway Administration
16. Abstract
Pronghorn (Antilocapra americana) movements were investigated with Global Position System (GPS)
telemetry from 2007 to 2008 along a 28-mile stretch of U.S. Highway 89 in northern Arizona to develop
strategies to enhance permeability with future highway reconstruction. Research objectives were to:
• Assess pronghorn movement patterns and distribution and determine highway permeability.
• Investigate the relationships of pronghorn distribution patterns to vehicular traffic volume.
• Assess the influence of fencing on pronghorn highway crossing patterns and permeability.
• Investigate pronghorn-vehicle collision patterns.
• Develop recommendations to enhance pronghorn highway permeability.
The team tracked 37 pronghorn (20 females, 17 males) with GPS receiver collars. Of 118,181 GPS fixes,
1,125 occurred within 0.15 mile of US 89, and 3,794 occurred within 0.30 mile. Only one pronghorn crossed
US 89 during the two years of tracking. The mean passage rate was a negligible 0.006 crossings/approach.
No collisions with vehicles were recorded during the study. In total, 5,035 weighted pronghorn approaches,
number of animals/segment, and five other criteria were used to rate 0.6-mile highway segments for
suitability as passage structure locations. The team recommended 3.2-mile spacing between passage
structures and three sites for passage structures integrated with fencing and noise reduction measures.
17. Key Words
Antilocapra americana, antelope, GPS telemetry,
fencing, highway impact, overpass, permeability,
pronghorn, traffic volume, wildlife passage structures,
wildlife-vehicle collisions
18. Distribution Statement
Document is available to the U.S.
public through the National
Technical Information
Service, Springfield, Virginia 22161
19. Security Classification
Unclassified
20. Security Classification
Unclassified
21. No. of Pages
81
22. Price
23.
Registrant’s
Seal
SI* (MODERN METRIC) CONVERSION FACTORS
APPROXIMATE CONVERSIONS TO SI UNITS APPROXIMATE CONVERSIONS FROM SI UNITS
Symbol When You Know Multiply By To Find Symbol Symbol When You Know Multiply By To Find Symbol
LENGTH LENGTH
in inches 25.4 millimeters mm mm millimeters 0.039 inches in
ft feet 0.305 meters m m meters 3.28 feet ft
yd yards 0.914 meters m m meters 1.09 yards yd
mi miles 1.61 kilometers km km kilometers 0.621 miles mi
AREA AREA
in2 square inches 645.2 square millimeters mm2 mm2 Square millimeters 0.0016 square inches in2
ft2 square feet 0.093 square meters m2 m2 Square meters 10.764 square feet ft2
yd2 square yards 0.836 square meters m2 m2 Square meters 1.195 square yards yd2
ac acres 0.405 hectares ha ha hectares 2.47 acres ac
mi2 square miles 2.59 square kilometers km2 km2 Square kilometers 0.386 square miles mi2
VOLUME VOLUME
fl oz fluid ounces 29.57 milliliters mL mL milliliters 0.034 fluid ounces fl oz
gal gallons 3.785 liters L L liters 0.264 gallons gal
ft3 cubic feet 0.028 cubic meters m3 m3 Cubic meters 35.315 cubic feet ft3
yd3 cubic yards 0.765 cubic meters m3 m3 Cubic meters 1.308 cubic yards yd3
NOTE: Volumes greater than 1000L shall be shown in m3.
MASS MASS
oz ounces 28.35 grams g g grams 0.035 ounces oz
lb pounds 0.454 kilograms kg kg kilograms 2.205 pounds lb
T short tons (2000lb) 0.907 megagrams
(or “metric ton”)
mg
(or “t”)
mg megagrams
(or “metric ton”)
1.102 short tons (2000lb) T
TEMPERATURE (exact) TEMPERATURE (exact)
ºF Fahrenheit
temperature
5(F-32)/9
or (F-32)/1.8
Celsius temperature ºC ºC Celsius temperature 1.8C + 32 Fahrenheit
temperature
ºF
ILLUMINATION ILLUMINATION
fc foot candles 10.76 lux lx lx lux 0.0929 foot-candles fc
fl foot-Lamberts 3.426 candela/m2 cd/m2 cd/m2 candela/m2 0.2919 foot-Lamberts fl
FORCE AND PRESSURE OR STRESS FORCE AND PRESSURE OR STRESS
lbf poundforce 4.45 newtons N N newtons 0.225 poundforce lbf
lbf/in2 poundforce per
square inch
6.89 kilopascals kPa kPa kilopascals 0.145 poundforce per
square inch
lbf/in2
SI is the symbol for the International System of Units. Appropriate rounding should be made to comply with Section 4 of ASTM E380
TABLE OF CONTENTS
EXECUTIVE SUMMARY ...................................................................................................1
1.0 INTRODUCTION.......................................................................................................5
1.1 BACKGROUND................................................................................................5
1.1.1 Pronghorn and Highways...............................................................................6
1.2 RESEARCH JUSTIFICATION.........................................................................8
1.3 RESEARCH OBJECTIVES...............................................................................10
2.0 STUDY AREA............................................................................................................11
2.1 PHYSICAL SETTING.......................................................................................11
2.2 CLIMATE ..........................................................................................................14
2.3 VEGETATION ..................................................................................................14
2.4 PRONGHORN POPULATION.........................................................................14
2.5 TRAFFIC VOLUME ........................................................................................16
3.0 METHODS..................................................................................................................17
3.1 PRONGHORN CAPTURE AND GPS TELEMETRY.....................................17
3.2 GPS DATA ANALYSIS OF PRONGHORN MOVEMENTS .........................17
3.2.1 Calculation of Passage Rates .........................................................................17
3.2.2 Calculation of Approaches and Weighted Approaches .................................19
3.2.3 Determination of Linear Approach Distance along Highway .......................20
3.3 PRONGHORN MOVEMENTS AND FENCING REMOVAL........................20
3.4 TRAFFIC VOLUME AND PRONGHORN DISTRIBUTION.........................21
3.5 PRONGHORN-VEHICLE COLLISIONS ........................................................21
3.6 IDENTIFICATION OF PASSAGE STRUCTURE SITES ...............................21
4.0 RESULTS....................................................................................................................25
4.1 PRONGHORN MOVEMENTS, DISTRIBUTION, AND APPROACHES.....25
4.1.1 Pronghorn Movements and Distribution........................................................25
4.1.2 Pronghorn Highway Crossings and Permeability ..........................................25
4.1.3 Pronghorn Approaches...................................................................................26
4.1.4 Linear Approach Distance along Highway....................................................27
4.2 PRONGHORN MOVEMENTS AND FENCING REMOVAL........................30
4.3 TRAFFIC VOLUME AND PRONGHORN DISTRIBUTION.........................30
4.4 PRONGHORN-VEHICLE COLLISIONS ........................................................30
4.5 IDENTIFICATION OF PASSAGE STRUCTURE SITES ...............................32
5.0 DISCUSSION .............................................................................................................35
5.1 PRONGHORN PERMEABILITY.....................................................................35
5.2 TRAFFIC VOLUME AND PRONGHORN DISTRIBUTION.........................36
5.3 STRATEGIES TO PROMOTE PRONGHORN PERMEABILITY .................37
5.3.1 Number and Spacing of Passage Structures...................................................38
5.3.2 Locations and Priorities for Potential Passage Structures..............................40
5.3.3 Role of ROW Fencing and Options...............................................................41
5.3.4 Types of Passage Structures and Specific Design Criteria ............................45
6.0 CONCLUSIONS AND RECOMMENDATIONS......................................................49
6.1 PRONGHORN PERMEABILITY.....................................................................49
6.2 POTENTIAL PASSAGE STRUCTURE LOCATIONS AND SPACING .......49
6.3 IMPACT OF TRAFFIC AND NOISE...............................................................50
6.4 ROLE OF FENCING .........................................................................................51
6.5 PASSAGE STRUCTURE DESIGN CRITERIA...............................................51
6.6 MONITORING ..................................................................................................53
REFERENCES ......................................................................................................................55
APPENDIX A – SUITABILITY RATINGS FOR PASSAGE STRUCTURES...................63
APPENDIX B - CON/SPAN® OVERPASS COST ESTIMATE AND PLANS ..................67
LIST OF TABLES AND FIGURES
Table 1. Mean probabilities that any GPS-collared pronghorn (n = 31) found within
distance bands from the highway at varying traffic volumes. .......................................... 32
Figure 1. Location of the US 89 research study area in north central Arizona. .............. 12
Figure 2. Study area stretch of US 89, extending from MP 430.0 to MP 458.0. ............ 13
Figure 3. Characteristic juniper woodland (top) and shortgrass prairie/grasslands
(bottom) associated with the US 89 research study area .................................................. 15
Figure 4. Hourly traffic volume (vehicles/hr) by hour along US 89 from 2007 to 2008
as determined by an ATR installed in 2007...................................................................... 16
Figure 5. Helicopter capture of pronghorn by net gunning (top; note the net over the
pronghorn), blindfolded and GPS-collared female to which an ear tag is being applied
(center), and the marked pronghorn being released near the US 89 study area................ 18
Figure 6. Distribution of GPS fixes for 37 pronghorn accrued from 2007 to 2008
adjacent to US 89. ............................................................................................................. 26
Figure 7. Frequency distribution among 0.1-mile segments of approaches to within 0.3
mile of US 89 made by 18 pronghorn on the west side of the highway (top) and by 13
pronghorn on the east side (bottom). .................................................................................28
Figure 8. Frequency distribution among 0.1-mile segments of weighted approaches to
within 0.3 mile of US 89 made by 18 pronghorn on the west side of the highway (top)
and by 13 pronghorn on the east side (bottom)................................................................. 29
Figure 9. Combined frequency distribution among 0.1-mile segments of weighted
approaches to within 0.3 mile of US 89 made by 31 pronghorn on both sides of the
highway.............................................................................................................................31
Figure 10. Mean probabilities that GPS-collared pronghorn (n = 31) occurred within
each 330-ft distance band from the highway at varying traffic volumes: a) <100, b)
100−200, c) 200−300, d) 300−400, e) 400−500, f) 500−600 vehicles/hr. ....................... 33
Figure 11. Ratings of suitability for pronghorn passage structures based on 11 criteria
by 0.6-mile segment between US 89 mileposts 440.0 and 458.0. .....................................34
Figure 12. Comparison of GPS fix distributions for three representative pronghorn
adjacent to US 89. ............................................................................................................. 39
Figure 13. Aerial view (top) and enlarged oblique view (bottom) from GoogleEarth©
depicting the proposed Coconino NF - Antelope Hills pronghorn passage structure site
on US 89 between MP 440.6−441.1. ................................................................................ 42
Figure 14. Aerial view (top) and enlarged oblique view (bottom) from GoogleEarth©
depicting the potential Wupatki National Monument pronghorn passage structure sites
between US 89 MP 444.2 and 444.6................................................................................. 43
Figure 15. Aerial view (top) and enlarged oblique view (bottom) from GoogleEarth©
depicting the Babbitt Ranch site between US 89 MP 447.2 and 447.7. ........................... 44
Figure 16. Renderings of potential pronghorn passage structures that emphasize
openness and unobstructed views for crossing pronghorn, including an overpass
capitalizing on existing terrain (top) and an elevated roadway/viaduct over gentle
terrain (bottom). .................................................................................................................47
Figure 17. Rendering of a CON/SPAN® pre-cast concrete arch application for a
wildlife overpass maintaining the integrity of a ridgeline. ................................................48
ACRONYMS AND ABBREVIATIONS
2-D two dimensional
3-D three dimensional
AADT average annual daily traffic
ADOT Arizona Department of Transportation
AGFD Arizona Game and Fish Department
ANOVA analysis of variance
ATR automatic traffic recorder
DPS Department of Public Service
EA environmental assessment
EPG Environmental Planning Group
ft foot/feet
FHWA Federal Highway Administration
GIS Geographic Information System
GMU Game Management Unit
GPS Global Positioning System
HR home range
hr hour
IGA intergovernmental agreement
in inch(es)
MCP minimum convex polygon
min minute(s)
mph miles per hour
NF National Forest
NPS National Park Service
ROW right(s)-of-way
SDI Shannon diversity index
SE standard error
SR State Route
SR 260 State Route 260
US 89 U.S. Highway 89
US 180 U.S. Highway 180
USFS U.S. Forest Service
VHF very high frequency
WCTAC Wildlife Connectivity Technical Advisory Committee
LIST OF SPECIES
Animals
Desert bighorn sheep Ovis canadensis
Caribou Rangifer tarandus
Elk Cervus elaphus
Grizzly bear Ursus arctos
Moose Alces alces
Mule deer Odocoileus hemionus
Pronghorn Antilocapra americana
White-tailed deer Odocoileus virginianus couesi
Wolf Canis lupus
Plants
Alkali sacaton Sporobolus airoides
Apache plume Fallugia paradoxa
Black grama Bouteloua eriopoda
Blackbrush Coleogyne ramosissima
Blue grama Bouteloua gracilis
Cliffrose Cowania mexicana
Galleta Pleuraphis jamesii
Greasewood Sarcobatus vermiculatus
Indian ricegrass Achnatherum hymenoides
Needle and thread grass Hesperostipa comata
One-seed juniper Juniperus monosperma
Ponderosa pine Pinus ponderosa
Pinyon Pinus edulis
Rabbit brush Ericameria nauseosa
Sagebrush Artemesia spp.
Saltbush Atriplex spp.
Shadscale Atriplex confertifolia
Winterfat Ceratoides lanata
ACKNOWLEDGEMENTS
This project was funded by the Arizona Department of Transportation’s (ADOT) Arizona
Transportation Research Center (ATRC), and the Federal Aid Wildlife in Restoration Act
Project W-78-R supporting Arizona Game and Fish Department (AGFD) research. The
research team commends ADOT for its proactive commitment to promoting wildlife
connectivity. The support of the Federal Highway Administration (FHWA), especially
Steve Thomas was also instrumental to the funding and conduct of the project.
Many individuals at ADOT provided endless support and guidance. Foremost was
Environmental Planning Group’s Justin White with his vision and commitment to
addressing permeability needs of pronghorn in the environmental analysis for U.S.
Highway 89 (US 89), and pursuing funding to make the project a reality. Estomih
Kombe of ATRC provided project oversight and coordination. The research team thanks
John Harper and Chuck Howe of the Flagstaff District for their tremendous support, as
well as commitment to implementing project recommendations. Doug Eberline of the
Transportation Planning Division provided traffic data support. The researchers
appreciate the support of Todd Williams, Bruce Eilerts, and Siobhan Nordhaugen of
Office of Environmental Services.
The active role of both the National Park Service (NPS) and US Forest Service (USFS) in
this project was greatly appreciated. The project would not have been possible without
their interest and commitment, as well as their proactive consideration of measures to
promote wildlife permeability. The research team sincerely thanks Diane Chung, Mary
Blasing, Steve Mitchelson, Paul Whitefield, Bob Van Belle and Brandon Holton of the
NPS, and Gene Waldrip, Cary Thompson, and Henry Provencio of the USFS for their
project support and facilitation.
AGFD Flagstaff Region personnel played a crucial role in supporting the project,
including Ron Sieg, Tom McCall, Carl Lutch, and Rick Miller. The outstanding capture
support provided by Larry Phoenix was vital to the success of the project. We are also
indebted to the capable pilots of Papillion Helicopters. The researchers thank Rob
Nelson and Chad Loberger for field project support.
Billy Cordasco of Babbitt Ranches was a tireless partner in this research project and
various pronghorn management activities adjacent to US 89. His ideas and suggestions,
pursuit of innovative fence treatments, and commitment as a partner in pursuing future
measures to promote pronghorn permeability are greatly appreciated.
The numerous volunteers associated with the Arizona Antelope Foundation were crucial
to the success of the pronghorn capture efforts. Further, their interest and commitment to
efforts to promote pronghorn permeability are noteworthy and sincerely appreciated.
Lastly, the Technical Advisory Committee for this project provided many suggestions
toward improving its effectiveness and applicability, and the research team appreciates
their guidance.
1
EXECUTIVE SUMMARY
Of all North American ungulate species, the barrier effect associated with highways
appears to affect no species as much as it does pronghorn antelope. The fragmentation of
pronghorn herds by highways has contributed to isolation of populations and disruption
of seasonal migrations, contributing to reduced pronghorn populations. Several previous
telemetry studies in northern Arizona, including adjacent to U.S. Highway 89 (US 89)
have demonstrated paved highways with fenced rights-of-way (ROW) constitute near
total barriers to pronghorn passage. While passage structures have proven effective for
other wildlife species, their application to promote pronghorn highway permeability has
been limited. The goal of this research project was to apply insights gained from
pronghorn movements and crossings of US 89 to develop strategies to enhance
permeability as part of future highway reconstruction. The specific objectives of this
project along were to:
1) Assess pronghorn movement patterns and distribution and determine current
permeability across the highway corridor.
2) Investigate the relationships of pronghorn highway crossing and distribution
patterns to vehicular traffic volume.
3) Assess the influence of fencing on pronghorn highway crossing patterns and
permeability.
4) Investigate pronghorn-vehicle collision patterns.
5) Develop recommendations to enhance pronghorn highway permeability.
The research team instrumented and tracked 37 pronghorn (20 females, 17 males) with
Global Positioning System (GPS) receiver collars from January 2007 to December 2008
along 28 miles of US 89; 19 pronghorn were captured on the west side and 18 on the east
side of the highway. Of the 118,181 GPS fixes accrued, 1,125 (1.0%) occurred within
0.15 mile of US 89, and 3,794 (3.2%) occurred within 0.30 mile of the highway, the
distance used to determine highway approaches and potential locations for passage
structures. During the GPS tracking, the pronghorn (n = 37) travelled an average of 3.2
miles each day. Most of the pronghorn (n = 31) were recorded within 0.30 mile of the
highway along a mean linear distance of 6.5 miles adjacent to US 89.
A single GPS-collared pronghorn crossed US 89 during the nearly two years of tracking;
none of the others did. The mean pronghorn crossing rate averaged 0.001 crossings/day
among the 30 animals that approached US 89 to within 0.15 mile. The mean pronghorn
passage rate was a negligible 0.006 crossings/approach. Due to the barrier effect and few
crossings by pronghorn, no collisions with vehicles were recorded during the study, nor
were any pronghorn-vehicle collision records found in ADOT’s roadkill database dating
back to 1990.
The frequency of approaches to within 0.30 mile of US 89 yielded considerably more
information than crossings to assist in the determination of potential pronghorn passage
2
structure locations. On the west side of US 89, 18 pronghorn approached the highway
2,875 times, for a mean of 159.7 approaches/animal. On the east side, 13 pronghorn
approached the highway 952 times, with a mean of 73.9 approaches/pronghorn. The
observed distribution of approaches from the east was not random. The research team
calculated weighted pronghorn approaches that accounted for the number of approaches,
number of different approaching animals, and the evenness of approaches over 0.1-mile
segments. Combined weighted approaches by pronghorn from both sides of the highway
totaled 5,035 approaches (16.2/segment). A significant peak accounting for nearly half
(47%) the approaches occurred on the highway section at the north end of the Coconino
National Forest (NF), which represents only 7% of the length of the area. Of the 31
pronghorn that approached the highway, 22 (71%) approached it in this 2-mile zone.
Pronghorn distribution remained constant among all distances and across all traffic
volumes up to 500 vehicles/hr. Only at volumes above 500 vehicles/hr was a change in
distribution observed. Pronghorn were consistently negatively impacted by traffic at
even low levels. Daytime traffic volumes along US 89 typically exceed the 10,000
vehicles/day level, the point at which highways become strong barriers to wildlife
passage and traffic repels animals away from the roadway. Pronghorn are primarily
active during daytime hours when peak traffic volumes occur along US 89.
At and adjacent to the 0.1-mile segment where barbed-wire fencing was removed from
the ROW fence within Wupatki National Monument approximately five years ago, there
was no evidence of any attempt to cross the highway. As such, it does not appear
pronghorn responded to the removal or modification of this short section of fencing.
The research team used pronghorn distribution and approaches in conjunction with five
other criteria to rate 0.6-mile segments for suitability as passage structure locations. The
research team recommended a spacing of 3.2 miles between passage structures. Based on
the rating criteria, three sites between mileposts (MP) 440.0 and 458.0 were
recommended as potential locations of passage structures. The most suitable location
was the section between MP 440.6 and 441.1, at the north end of the Coconino NF.
Another recommended site is on Wupatki National Monument (MP 444.2−444.6) three
miles to the north of the aforementioned site. This site is attractive due to the ease of
addressing ROW fencing issues (since no livestock grazing occurs here), the presence of
high-quality pronghorn habitat, and the reconstructed highway’s planned median width,
which is considerably narrower than that of other recommended sites. A third recom-mended
passage structure location is on Babbitt Ranch lands at MP 447.2−447.7
and spans both private and State Trust land. The high rating for this segment reflects
Babbitt Ranches’ proactive role in pronghorn management, including further
modification of ROW fencing.
No passage structure designed specifically to accommodate pronghorn passage has been
constructed in North America. As such, limited guidelines or insights exist as to what
types of structure are best suited to promoting pronghorn permeability. The research
team believes that overpasses and/or large elevated viaducts have the best potential for
promoting permeability along US 89. Site specific characteristics associated with the
3
different passage structure locations will dictate what type of potential structure might be
appropriate from engineering and cost standpoints. The most important structural
consideration is the requirement that passage structures be as open and wide as possible,
with attention paid to avoiding obstructed line-of-sight views through or across
structures.
The terrain near MP 441 is suited to the construction of an overpass, and since this stretch
of US 89 has been previously reconstructed, a retrofit application is appropriate. The
application of a pre-cast concrete arch overpass may hold potential. The research team
recommends that implementation of such a structure at MP 440.9 be considered under an
experimental enhancement grant. Insights would be gained on the efficacy of a passage
structure where the prospect for success is highest. Further, the estimated cost for the
structure not including fill material ($542,725 by one estimate) and relative ease of
construction (just a few days) for an overpass makes an enhancement grant approval a
possibility.
At the Wupatki National Monument site, the variation in terrain will support either an
overpass near a ridgeline or an elevated viaduct. At the Babbitt Ranch site where terrain
is predominantly flat, an elevated viaduct would function best in promoting permeability.
Ideally, passage structures should be located in areas with no ROW or livestock pasture
fencing near the highway such that the fencing presents an impediment to free passage by
pronghorn. Where it cannot be avoided, creative approaches should be used to minimize
the barrier effect of fencing near passage structures. A comprehensive set of measures
should be used to create “quiet zones” around passage structures to facilitate pronghorn
highway approaches and crossings by reducing traffic-associated noise’s impact. Such
measures include recessing the roadway below grade, integrating noise barriers, planting
vegetation, erecting sound walls, and applying pavement treatments like rubberized
asphalt.
This project reflects an incremental and proactive process of addressing permeability and
habitat continuity for pronghorn along US 89. The project reflects ADOT’s commitment
to obtaining data to make informed data-driven decisions in highway reconstruction
planning on the need of and best locations for passage structures to promote pronghorn
permeability.
4
5
1.0 INTRODUCTION
1.1 BACKGROUND
Direct and indirect highway impacts have been characterized as some of the most
prevalent and widespread forces altering natural ecosystems in the U.S. (Noss and
Cooperrider 1994, Trombulak and Frissell 2000, Farrell et al. 2002). Forman and
Alexander (1998) and Forman (2000) estimated that highways have affected more than
20% of the U.S. land area through habitat loss and degradation. Mortality from vehicle
collisions is a serious and growing problem for wildlife populations, and also contributes
to human injuries, deaths, and tremendous property loss (Reed et al. 1982, Farrell et al.
2002, Schwabe and Schuhmann 2002, Bissonette and Cramer 2008). An even more
pervasive impact of highways on wildlife is the indirect barrier and fragmentation effects
resulting in diminished habitat connectivity and permeability (Noss and Cooperrider 1994,
Forman and Alexander 1998, Forman 2000, Forman et al. 2003, Bissonette and Adair
2008). Highways act as barriers to free movement of wildlife between seasonal ranges or
other vital habitats (Trombulak and Frissell 2000). Highways fragment and isolate habitats
and populations, limit juvenile dispersal (Beier 1995), and reduce genetic interchange
(Epps et al. 2005, Riley et al. 2006), all serving to disrupt viable wildlife population
processes. Long-term fragmentation and isolation increases population susceptibility to
random catastrophic events (Swihart and Slade 1984, Forman and Alexander 1998,
Trombulak and Frissell 2000).
While many studies have alluded to highway barrier effects on wildlife (e.g., Forman et
al. 2003), few have yielded quantitative data to measure permeability or quantify the
barrier effect, particularly in an experimental (e.g., before and after construction) context
with research controls (Hardy et al. 2003, Roedenbeck et al. 2007, Dodd et al. 2007a,
Olsson 2007). Many studies have focused on the efficacy of passage structures in
promoting passage (Clevenger and Waltho 2003, Ng et al. 2004). Dodd et al. (2007a)
stressed the value of a quantifiable metric of permeability and calculated elk highway
passage rates from GPS telemetry to conduct before-after-control reconstruction
comparisons along State Route (SR) 260. They found that overall elk passage rates
averaged 0.50 crossings/approach; among reconstruction classes, the mean elk passage
rate for the before-reconstruction/control class (0.67) was 39% higher than the mean
after-reconstruction passage rate (0.41). Dodd et al. (2009) also calculated white-tailed
deer passage rates along SR 260, which averaged only 0.03 crossings/approach on control
sections. Paquet and Callaghan (1996) reported that passage rates for wolves averaged
0.93 crossings/approach along a low-traffic highway but only 0.06 along the Trans-
Canada Highway. Waller and Servheen (2005) compared grizzly bear highway crossing
frequency determined by GPS telemetry to simulated random walk analyses to assess
permeability; observed crossing frequency was 31% of the simulated frequency. Dyer et
al. (2002) compared actual road crossings to simulated crossing rates. They found that
caribou road crossings were 20% lower than suggested by the simulations. Olsson (2007)
documented an 89% decrease in the mean moose-crossing rate between before- and after-reconstruction
levels along a highway in Sweden.
6
1.1.1 Pronghorn and Highways
Highways’ barrier effect appears to consistently affect pronghorn antelope more than any
other North American ungulates. The fragmentation of pronghorn herds by highways,
railways, canals, fences, human encroachment, and habitat degradation has contributed to
isolation of populations and disruption of seasonal migrations, thereby contributing to a
reduction of pronghorn populations (O’Gara and Yoakum 1992, Sawyer and Rudd 2005).
Pronghorn are a nomadic species capable of long-distance movements in response to
extreme seasonal weather conditions and variable forage and water availability (Yoakum
and O’Gara 2000). Historically, pronghorn roamed freely in North America including
northern Arizona (Yoakum and O’Gara 2000), but populations declined as much as 99%
by the early 1900s (Yoakum 1968). In Arizona, populations declined from
approximately 45,000 animals in the 1900s (Knipe 1944) to only 7,500 by 2002 (AGFD,
unpublished data), and have since increased through aggressive management to 11,000
(AGFD 2007).
As early as 1950, Buechner (1950) recognized that fenced highways restricted pronghorn
movement across Texas highways. In Wyoming, a state that harbors 60% of North
America’s pronghorn, Interstate-80 has long been considered a significant barrier to
pronghorn movement (Sheldon 2005). Several VHF telemetry studies in northern
Arizona have demonstrated that paved highways with fenced rights-of-way (ROW)
constitute near total barriers to pronghorn passage. Ockenfels et al. (1994) tracked 47
animals adjacent to Interstate-17 and found that no individual pronghorn was observed on
both sides of the highway. None had crossed the highway. Likewise, Ockenfels et al.
(1997), van Riper and Ockenfels (1998), and Bright and van Riper (2000) never
documented any pronghorn crossings of the fenced highways they monitored: US 89 at
Wupatki National Monument, Interstate-40, U.S. Highway 180 (US 180), or a railroad at
Petrified Forest National Park. Ockenfels et al. (1997) and van Riper and Ockenfels
(1998) did however report that pronghorn crossed the low traffic-volume, paved but
unfenced park road through Wupatki National Monument. Hart et al. (2008) confirmed
that the railroad remained a total barrier to passage of eight collared pronghorn, even
after the fence next to the railroad was modified to promote passage. These Arizona
studies point to the combined impact of fenced ROW and highways with associated
traffic, though it is difficult to partition their contributory impact to reduced pronghorn
permeability. Sheldon (2005) found that fences in Wyoming significantly influenced
pronghorn movements and distribution, and that home ranges were located in areas with
the lowest fence densities. The presence and type of ROW fences determined whether
roads were included in seasonal ranges and where pronghorn crossed highways. Sheldon
(2005) also found that seasonal crossings consistently occurred along unfenced highway
sections.
Limited information exists on the relationship of highway traffic volume to pronghorn
movement and distribution patterns; such information could help assess the relative
impact attributable to highways and fences. Theoretical models (Mueller and Berthoud
1997) suggest that highways averaging 4,000−10,000 vehicles/day present strong barriers
to wildlife and would repel animals from the highway. Whereas most ungulate highway
7
crossings (e.g., elk and deer) occur during nighttime hours when traffic volume is lowest,
pronghorn are diurnal and active when traffic volumes are typically at their highest
(Gagnon et al. 2007a). Gagnon et al. (2007b) found that increasing vehicular traffic
volume decreased the probability of at-grade crossings by elk and that they moved away
from the highway, consistent with Mueller and Berthoud (1997). For white-tailed deer,
Dodd et al. (2009) found that at-grade passage rates were consistently low (≤0.1
crossings/approach) across all traffic volumes. Regular vehicular traffic on roads in
Wyoming was reported to produce minimal disturbance among pronghorn due to
habituation, though females with young remained sensitive to vehicular traffic (Reeve
1984). Gavin and Komers (2006) reported that pronghorn in Alberta exhibited higher
proportions of vigilant behavior along high traffic roads during spring compared to lower
traffic roads, suggesting that traffic volume influenced risk perception. Pronghorn close
to roadways exhibited higher vigilance regardless of traffic levels, further suggesting they
perceived roads to be a danger. Gavin and Komers (2006) also found that individuals in
pronghorn herds with young were more risk averse than other social groupings; this is
consistent with Reeve’s findings (1984).
Integration of structures designed to promote wildlife passage across highways in
transportation projects has increased in North America, particularly structures (e.g.,
underpasses or overpasses) designed specifically for large animal passage (Foster and
Humphrey 1995; Clevenger and Waltho 2003; Gordon and Anderson 2003; Dodd et al.
2007b, 2009). Wildlife passage structures have shown benefit in promoting wildlife
passage for a variety of wildlife species (Farrell et al. 2002; Clevenger and Waltho 2003;
Dodd et al. 2007b, 2009). Dodd et al. (2007c, 2009) found that elk passage rates along
one section of SR 260 increased 52% to 0.81 crossings/approach once reconstruction was
completed and ungulate-proof fencing linking passage structures was installed. This
pointed to the efficacy of passage structures and fencing in promoting permeability, as
well as achieving an 85% reduction in elk-vehicle collisions (Dodd et al. 2006). Gagnon
et al. (2007c) found that traffic levels did not influence elk passage rates during below-grade
underpass crossings. This finding shows the benefit of underpasses and fencing in
promoting permeability by funneling elk to underpasses where traffic has minimal effect
compared to crossing at-grade in areas with high traffic volumes (Gagnon et al. 2007b).
Dodd et al. (2009) reported five-fold higher white-tailed deer permeability (0.16
crossings/approach) along SR 260 after passage structures were added during
reconstruction than the control's (0.03); like elk, deer passage rates were minimally
affected by traffic on sections where passage structures facilitated below-grade passage.
While passage structures were shown to be effective in promoting below-grade crossings,
Dodd et al. (2009) found considerable variation in mean passage rates on three
reconstructed highway sections, ranging from 0.09 to 0.81 crossings/approach. This
likely reflected the corresponding variation in passage structure spacing ranging from 1.5
to 0.6 miles between structures; there was a strong inverse association (r = -0.847)
between passage rate and passage structure spacing. Bissonette and Adair (2008)
conducted an assessment of recommended passage structure spacing for several species
tied to isometric scaling of home ranges (HR). They used HR0.5 as a daily movement
metric and passage structure spacing distance, which when used with other criteria will
8
help maintain landscape permeability. Bissonette and Adair (2008) recommended
spacing of 2.0 miles between passage structures for pronghorn.
While passage structures have proven effective for other wildlife species, their
application to promote pronghorn highway permeability has been limited. (Sawyer and
Rudd 2005). Though Plumb et al. (2003) documented 70 crossings by pronghorn at a
concrete box-culvert underpass in Wyoming (81% in a single crossing), pronghorn
overall exhibited reluctance to use the structure and the majority of crossing pronghorn
accompanied mule deer through the underpass; crossing pronghorn comprised a small
proportion of the local pronghorn herd. In six years of monitoring underpasses along
Interstate-80 through which thousands of mule deer passed, only a single pronghorn was
recorded passing through the structures monitored by Ward et al. (1980). In spite of the
limited use of structures to date, there is recognition of the need for strategies to promote
pronghorn permeability (Ockenfels et al. 1994, Hacker 2002, Yoakum 2004, Sawyer and
Rudd 2005). Yoakum (2004) believed that pronghorn behavioral characteristics might
preclude effective use of both underpasses and overpasses on high-volume highways.
Sawyer and Rudd (2005:6) reported that “with the exception of Plumb et al. (2003) and
several anecdotal observations, we could not find any published or documented
information on pronghorn utilizing crossing structures.” Still, they believed that pronghorn
would more readily use open-span bridges as underpasses than they would use overpasses,
though no studies have been done to support this contention. To date, no passage structure
intended for pronghorn passage has been implemented in North America. Corlatti et al. (2009)
argued for long-term monitoring and genetic studies to evaluate passage structure effective-ness
in promoting population growth and genetic viability. They believed monitoring was
needed to justify the building of overpasses during highway projects as a means to maintain
connectivity, in view of their high cost. Such an argument is particularly relevant to prong-horn
since highways present such significant barriers and there is such a limited application of
passage structures and little insight on the benefits to promoting connectivity and gene flow.
1.2 RESEARCH JUSTIFICATION
US 89 is the primary highway route connecting Flagstaff/Interstate-40 with Utah to the
north, and serves the Navajo Nation and popular recreation areas north of Flagstaff (e.g.,
Sunset Crater and Wupatki national monuments, Grand Canyon National Park, Page,
Lake Powell, etc.). The final US 89 Antelope Hills – Junction US 160 Environmental
Assessment (ADOT 2006) addressed alternatives for the reconstruction of US 89. The
environmental assessment (EA) reported that traffic volume along US 89 (currently
averaging 7,500 vehicles/day; ADOT 2006) is projected to double in 20 years. The
majority of the existing highway is a 2-lane roadway with limited passing lanes. Under
the preferred alternative, US 89 would be widened to a 4-lane divided highway; the
center median along three miles through Wupatki National Monument would be 30 ft
wide, while from there to Gray Mountain the median would be 84 ft wide.
As documented in the EA, the primary environmental effect of the proposed US 89
reconstruction on pronghorn populations would be to increase the barrier effect
9
associated with the widened highway and increased traffic, contributing to a higher
degree of population fragmentation. It is recognized that a “wide, naturally vegetated
overpass structure over US 89 may facilitate pronghorn movement across the US 89
corridor” (ADOT 2006:76). The EA also addressed secondary impacts from highway
reconstruction on pronghorn, citing the loss of connectivity and genetic viability.
The EA states:
“ADOT in consultation with FHWA will make a good faith effort to find the
funding for a proposed 3-year research project to determine pronghorn
movements north of I-40 in Arizona that will be completed at a minimum 1 year
prior to final design for projects between milepost 442.0 to milepost 458.0
(Navajo Indian Reservation boundary). ADOT’s Environmental Planning Group
(EPG) will coordinate the pronghorn research project and will establish at the
beginning of the research project a Wildlife Connectivity Technical Advisory
Committee (WCTAC) consisting of representatives from FHWA, NPS, NFWD
(Navajo Fish and Wildlife Department), and AGFD. The NPS, as a cooperating
agency, has committed to maintain connectivity for pronghorn and other species
as part of its requirements under the NPS Organic Act (16 US Code 1–4) and NPS
policy. The WCTAC will review data from the research project, provide
recommendations to ADOT and FHWA on the appropriateness of a pronghorn
crossing structure, and identify the potential location and conceptual design of a
crossing structure, if warranted for consideration prior to final project design. The
WCTAC will also address wildlife connectivity in general for the project (ADOT
2006:76).”
In November 2004 (following issuance of the draft EA), EPG commissioned a research
concept paper to implement the research project addressed in the EA. In November
2006, the Arizona Department of Transportation (ADOT) and the Arizona Game and
Fish Department (AGFD) executed an Interagency Agreement between for the research
project (Project JPA07-004T) with funding provided by ADOT’s Arizona Transportation
Research Center. This research project is significant from several perspectives. First, it
epitomizes the incremental process in addressing wildlife connectivity and permeability
needs embodied in Arizona’s Wildlife Linkages Assessment (Arizona Wildlife Linkages
Workgroup 2006). General connectivity needs identified in the assessment (e.g., Linkage
No. 17; Deadman Mesa to Gray Mountain) were proactively addressed in the EA process
for US 89. This led to the commitment to obtain information to make data-driven
decisions on the need and best location(s) for passage structures to promote pronghorn
permeability that could be built during highway reconstruction. Further, compared to
previous ADOT-funded research on wildlife-highway relationships along SR 260, SR 64,
and Interstate-17, there was no overarching highway safety issue associated with wildlife-vehicle
collisions (Dodd et al. 2006, 2007a, 2009); rather, this research project was
predicated solely on addressing ecological needs for pronghorn connectivity.
10
1.3 RESEARCH OBJECTIVES
This research project will add greatly to the understanding of pronghorn movements in
relation to highways and traffic volume. Given the degree to which highways are known
to limit pronghorn permeability, the depressed nature of current pronghorn populations
and the fact that US 89 will be reconstructed in the future to accommodate increasing
traffic volume, the challenge is to determine if and how pronghorn meta-populations in
northern Arizona can be reconnected to maintain population viability. The overarching
goal of this research project was to apply insights gained on current pronghorn
movements and permeability across US 89 in developing strategies to enhance
connectivity in future highway reconstruction. The specific objectives of this research
project were to:
1) Assess pronghorn movement patterns and distribution relative to US 89 and
determine current permeability across the highway corridor.
2) Investigate the relationships of pronghorn highway crossing and distribution
patterns to vehicular traffic volume.
3) Assess the influence of fencing on pronghorn highway crossing patterns and
permeability.
4) Investigate pronghorn-vehicle collision patterns along US 89.
5) Establish a baseline to assess the degree to which US 89 and other northern
Arizona highways have affected gene flow and genetic diversity within and
among pronghorn populations.
6) Develop recommendations to enhance pronghorn highway permeability.
11
2.0 STUDY AREA
The focus of this research project was a 28-mile stretch of US 89 starting approximately
15 miles northeast of Flagstaff, Coconino County, Arizona (lat 35o22’–35o46’N, long
111o20’–111o40’W). The study section stretches from milepost (MP) 430.0 at the
northern end of the Coconino National Forest (NF) near the entrance to Sunset Crater
National Monument, to MP 458.0 at the Navajo Nation boundary north of Gray Mountain
(Figures 1 and 2). A 3-mile portion of the highway crosses through Wupatki National
Monument near the center of the study area (Figures 1 and 2). US 89 is classified as a
Rural Principal Arterial highway; these highways are considered the state’s principal
corridors for statewide travel— they carry the highest volume of long distance trips in
Arizona (ADOT 2006).
In 1999, US 89 from MP 430.0 to 442.0 was reconstructed to a 4-lane divided highway.
From MP 442.0 north, the section of US 89 proposed for reconstruction in the US 89
Antelope Hills – Junction US 160 Environmental Assessment (ADOT 2006) is
predominantly a 2-lane roadway with occasional passing lanes and lateral road access
turnouts. Within the study area, the planned design standards north of MP 442.0 include:
�� A 5-lane undivided roadway with two lanes in each direction and a two-way
continuous left-turn lane centered on the existing centerline in the Antelope Hills
section (MP 442.0 to MP 443.0).
• A 4-lane divided section with 30-foot median width, widening centered about the
existing centerline on the Wupatki National Monument section (MP 443.0 to MP
445.4).
• A 4-lane divided section with 84-foot median width, widening to the west of the
existing centerline with the exception of between MP 445.4 and MP 447.0 and
between MP 451.7 and MP 453.2, where the roadway will be widened to the east
to avoid impacting existing business for the Wupatki National Monument to Gray
Mountain section (MP 445.4 to MP 456.0).
Land ownership adjacent to the highway corridor includes U.S. Forest Service (USFS)
(Coconino NF) lands on the south half interspersed with scattered small private land
parcels, National Park Service (NPS) (Wupatki National Monument) lands in the center
of the study area, and a “checkerboard” pattern of Arizona State Trust and private land
(primarily Babbitt Ranch) holdings on the north half of the study area (Figures 1 and 2).
2.1 PHYSICAL SETTING
The study area is located at the southwestern extent of the Colorado Plateau
physiographic province, and lies within the San Francisco Peaks Volcanic Field (Hansen
et al. 2004). The study corridor lies adjacent to US 89. One end is at 6,890 ft elevation
near the Sunset Crater National Monument turnoff atop Deadman Mesa on the east flank
of the San Francisco Peaks. It steadily drops off to the north along Deadman Flats and
12
Figure 1. Location of the US 89 research study area in north central Arizona (map from
Hansen et al. 2004).
continues to Gray Mountain at the northern extent of the study area to an elevation of
4,900 ft. The geology and topography within the study area are a diverse and complex
mix of mesas, basalt flows, cinder cones extending northward through much of the study
area (Figure 2), rolling hills, and arroyos, all interspersed with relatively flat grassland
areas. At the northern extent lie broken bluffs and sparsely vegetated badlands associated
with the Painted Desert. The eastern extent is defined by cliffs and bluffs above the Little
Colorado River and the broken lowlands of the Wupatki Basin (Hansen et al. 2004). This
diversity in elevation and topography has a significant influence on vegetative
community composition (Hansen et al. 2004), which in turn influences pronghorn
distribution and habitat use (Ockenfels et al. 1997, Bright and van Riper 2000). The
numerous manmade stock tanks and some springs that are scattered throughout the area
also influence pronghorn distribution (Bright and van Riper 2000).
Research
Study Area
13
Figure 2. Study area stretch of US 89, extending from MP 430.0 to 458.0
.
14
2.2 CLIMATE
The variation in elevation and topography across the study area affects climatic patterns.
Most of the study area is semi-arid, dominated by hot summers and cool winters. At the
lower elevations of the area’s northern part, precipitation is low and annually averages
only 5.2 in, with occasional winter snows that annually average 8.9 in (van Riper and
Ockenfels 1998). Summer thunderstorms account for the majority of precipitation in the
northern portion of the area (Hansen et al. 2004). Here, summer temperatures often
exceed 100o F and winter lows typically hover around freezing but can occasionally dip
to 10oF after winter storms. At the southern end of the study area, with higher elevations
and the nearby San Francisco Peaks, precipitation is considerably higher and more
consistent, averaging 19.8 in, with considerable snowpack accumulating during winter
(van Riper and Ockenfels 1998). Due to the presence of the San Francisco Peaks south
of the study area, windy conditions often prevail which further exert an influence on
pronghorn distribution and habitat use.
2.3 VEGETATION
Vegetation within the study area is diverse and exhibits characteristics of the Montane
Coniferous, Plains, Great Basin Grassland, and Great Basin Desertscrub biotic
communities (Brown 1994, Hansen et al. 2004). Dominant plant species in the southern
portion include a ponderosa pine and limited pinyon pine overstory with sagebrush,
rabbit brush, cliffrose, and Apache plume in the understory, interspersed with small
grasslands composed primarily of blue grama and other grasses. At lower elevations, the
vegetation is dominated by oneseed juniper woodlands with cliffrose, Apache plume, and
other shrubs, along with blue grama and other grasses (Figure 3). Juniper woodlands
transition to shortgrass prairie/grasslands composed of blue and black grama, galleta,
alkali sacaton, and needle and thread grasses, with winterfat and sagebrush interspersed with
sparse junipers (Hansen et al. 2004; Figure 3). At the northern extent of the area, desertscrub
vegetation is dominated by shadscale, greasewood, rabbit brush, and blackbrush, with
Indian ricegrass.
Most of the study area has a long history of livestock grazing, which has altered plant
communities, particularly grasslands, and contributed to juniper encroachment (ADOT
2006, Hansen et al. 2004). Though grazed by livestock until the 1980s, Wupatki
National Monument supports relatively pristine native bunchgrass grasslands that provide
reference conditions for historical grasslands and offer a seed source for dispersal to
surrounding habitats (ADOT 2006). Wupatki National Monument constitutes excellent
pronghorn habitat for forage and especially cover, particularly when adjacent national
forest, state trust and private lands are grazed by livestock (Bright and van Riper 2000; B.
Holton, unpublished report, NPS, Flagstaff, AZ).
2.4 PRONGHORN POPULATION
Two distinct pronghorn herds inhabit the study area, one on each side of US 89 (ADOT
2006). Both herds fall within AGFD Game Management Unit (GMU) 7, and as such are
surveyed and managed as a single population. However, based on movement studies by
15
Figure 3. Characteristic juniper woodland (top) and shortgrass prairie/grasslands
(bottom) associated with the US 89 research study area (photos from Hansen et al. 2004).
Ockenfels et al. (1997), van Riper and Ockenfels (1998), and Bright and van Riper
(2000), who documented limited passage across US 89, these herds have become
virtually separate and isolated. The herd on the west side of US 89 ranges westward to
US 180. The herd to the east ranges to the Little Colorado River and south to Interstate-
40. The current (2007−2008) population estimate for GMU 7 is approximately 600
16
animals (AGFD unpublished Pronghorn Hunt Recommendations, Game Branch,
Phoenix). From 2003 to 2008 surveys of the population from fixed wing aircraft found
an average of 220 pronghorn in 42 groups with an average ratio of 36 males (bucks):100
females (does):40 young (fawns). The general population trend has been downward and
is a source of concern, particularly given the large numbers of animals (males) harvested
by sport hunting (AGFD unpublished Pronghorn Hunt Recommendations, Game Branch,
Phoenix). Since 2003, the fawns:doe ratio on the west side of US 89 has averaged 0.45
compared to only 0.27 on the east side.
2.5 TRAFFIC VOLUME
Average annual daily traffic (AADT) volume on this portion of US 89 (sampled at Gray
Mountain) was estimated at 5,600 vehicles/day in 2006 and 7,300 in 2007 (unpublished
data, ADOT Data Management Section). Since March 2007, traffic volume has been
continuously measured by a permanent automatic traffic recorder (ATR) installed near
the center of the study area just north of Wupatki National Monument. This ATR
measured an actual AADT of 6,310 vehicles/day in 2008. Traffic volumes were highest
during daytime hours (Figure 4). Between 10:00 and 17:00, hourly traffic volume
exceeded 430 vehicles/hr, equivalent to a volume of 10,000−11,600 vehicles/day.
Monthly traffic volume was highest during May−August when it averaged 60% higher
than volume during the lowest traffic months of December−February. Passenger cars
accounted for 81% of all vehicles traveling along US 89 (2007-2008), though commercial
trucks accounted for up to a third of the traffic during early morning hours (midnight to
03:00). Vehicular speeds averaged 72.5 mph though the posted speed at the ATR and
along most of the study area stretch of US 89 is 65 mph.
Figure 4. Hourly traffic volume (vehicles/hr) by hour along US 89, Arizona from 2007
to 2008, determined by an automatic traffic recorder installed in 2007.
17
3.0 METHODS
3.1 PRONGHORN CAPTURE AND GPS TELEMETRY
The research team captured pronghorn using a net gun fired from a helicopter (Firchow et
al. 1986, Ockenfels et al. 1994; Figure 5). A fixed-wing aircraft and numerous ground
spotters using optics equipment were employed to search for pronghorn during capture to
minimize helicopter searching. Pronghorn were primarily captured during the winter
(December−January) to minimize heat-related stress on animals, as well as deleterious
effects on females that could occur if captured later in their pregnancies. The team’s
capture objectives were to: 1) instrument as nearly an equal number of pronghorn on each
side of US 89 as possible, 2) spread the collars among as many different herds along the
length of the study area as possible, and 3) capture animals within five miles of US 89.
Upon capture, pronghorn were immediately blindfolded and untangled from the capture
net. Animals were fitted with a GPS collar and marked with a numbered, colored ear tag
(Figure 5). Tissue samples were taken from the animals’ ears with a paper punch and
preserved for future genetic analysis. The research team instrumented the pronghorn
with store-on-board GPS receiver collars (Model TGW-3500; Telonics, Inc., Mesa, AZ)
programmed to receive 12 GPS fixes/day, with one fix every 90 min between
04:00−22:00; the GPS units had a battery life of 11 months. All collars had VHF
beacons, mortality sensors, and programmed release mechanisms to allow recovery.
3.2 GPS DATA ANALYSIS OF PRONGHORN MOVEMENTS
Once the GPS collars were recovered and data downloaded, the research team employed
ArcGIS Version 8.3 Geographic Information System (GIS) software (ESRI, Redlands,
California) to analyze the data similar to analyses done for elk by Dodd et al. (2007d,
2009) and white-tailed deer (Dodd et al. 2009). The team used GPS data to calculate
daily distance traveled by the collared pronghorn by sex and season, as well as individual
minimum convex polygon1 (MCP) home ranges comprised of all GPS fixes (White and
Garrott 1990). Differences in means were assessed by analysis of variance (ANOVA),
and means were reported with ±1 standard error (SE).
3.2.1 Calculation of Passage Rates
The team divided the study length of US 89 into 280 sequentially numbered 0.1-mile
segments corresponding to the units used by ADOT for tracking wildlife-vehicle
collisions and highway maintenance, and identical to Dodd et al. (2007d, 2009). The
number and proportion of GPS pronghorn fixes within 0.15, 0.30, and 0.60 mile of US
89 were calculated for each animal, as well as the proportion of three-dimensional (3-D)
or two-dimensional (2-D) fixes that were acquired.
1 Constructed by connecting the outermost fixes.
18
Figure 5. Helicopter capture of pronghorn by net gunning (top; note the net over the
pronghorn), blindfolded and GPS-collared female to which an ear tag is being applied
(center), and the marked pronghorn being released near the US 89 study area (R.
Ockenfels photos).
19
The team drew lines connecting all consecutive GPS fixes and inferred a highway
crossing where lines between fixes crossed the highway through a given segment (Dodd et
al. 2007d, 2009). Animal Movement ArcView Extension Version 1.1 software (Hooge
and Eichenlaub 1997) was used to assist in determining where pronghorn had crossed. The
research team compiled crossings by individual animals by highway segment, date and
time, and calculated crossing rates for individual pronghorn by dividing the number of
crossings by the days a collar was worn. As stated earlier, it turned out there was only
one crossing detected.
Passage rates for individual collared pronghorn were used as the relative measures of
highway permeability (Dodd et al. 2007d, 2009). An approach was considered to have
occurred when an animal traveled from a point outside the 0.15-mile buffer zone to a
point within 0.15 mile of US 89, determined by successive GPS fixes. The approach
zone corresponded to the road-effect zone associated with traffic-related disturbance
(Rost and Bailey 1979, Forman et al. 2003) previously used for elk and white-tailed deer
by Dodd et al. (2007d, 2009). Pronghorn that directly crossed US 89 from a point
beyond 0.15 mile were counted as an approach and a crossing.
3.2.2 Calculation of Approaches and Weighted Approaches
Based on previous pronghorn telemetry research adjacent to US 89 (Ockenfels et al.
1997, van Riper and Ockenfels 1998, Bright and van Riper 2000), the research team
anticipated that there might be few pronghorn crossings or approaches to within 0.15
mile, especially when compared to the 11,052 crossings by 100 elk along SR 260 (Dodd
et al. 2009). As such, the team used the number of approaches by pronghorn to within
0.30 mile to determine the distribution of animals adjacent to US 89 for the purposes of
assessing the need for and potential location(s) of passage structures. Use of this greater
approach distance also was deemed appropriate given the relatively open nature of
pronghorn habitat, pronghorn reliance on visual stimuli in risk avoidance (Gavin and
Komers 2006), and pronghorn mobility over long distances compared to other ungulates
(Yoakum and O’Gara 2000).
To account for the number of individual pronghorn that approached each highway
segment adjacent to US 89, as well as evenness in crossing frequency among animals, the
research team calculated Shannon diversity indices (SDI; Shannon and Weaver 1949) for
each segment using this formula:
Thus, to calculate SDI (or H′ ) for each highway segment, the researchers calculated and
summed all the -(pi ln pi) for each pronghorn that had approaches in the segment, where
each pi is defined as the number of individual collared pronghorn approaches within each
segment divided by the total number of pronghorn approaches in the segment. SDI were
used to calculate weighted approach frequency estimates for each highway segment,
multiplying uncorrected approach frequency × SDI. Weighted approaches better
20
reflected animal approach frequency, number of approaching animals, and equity in
distribution among approaching pronghorn (cf. Dodd et al. 2006, 2007a).
Pronghorn highway approaches were determined for animals approaching from each side
of US 89, and both sides combined. The research team tested the hypothesis that the
observed spatial approach distribution (by 0.10-mile segments) did not differ from a
discrete randomly generated approach distribution using a Kolmogorov-Smirnov test
(Clevenger et al. 2001; Dodd et al. 2006, 2007d), a test that is sensitive to both the
difference in ranks and shape of the distributions.
3.2.3 Determination of Linear Approach Distance along Highway
To assist with the assessment of the number and spacing of passage structures that might
be necessary to promote pronghorn passage across US 89, the research team compiled the
linear distance adjacent to US 89 between the 0.1-mile segments in which pronghorn
approached that were the furthest apart. This linear approach distance measured how far
animals ranged along the length of US 89; for example, an animal that had approaches
within segments 21 through 175, spanning 154 0.1-mile segments, had a linear approach
distance of 15.4 miles.
3.3 PRONGHORN MOVEMENTS AND FENCING REMOVAL
Around 2004, barbed wire was removed from a short 0.1-mi section of the ROW fence
between MP 444.1 to MP 444.2 on Wupatki National Monument to facilitate pronghorn
crossing, though the fence T-posts were left in place. As detailed in the US 89 Antelope
Hills – Junction US 160 Environmental Assessment (ADOT 2006), there is no conclusive
evidence that pronghorn have crossed at this point aside from a few pronghorn tracks
being found on both sides of the highway in this location. As part of its pronghorn GPS
telemetry tracking, the research team assessed whether animals approached and/or
crossed US 89 at this point to a higher degree than the adjacent sections of highway.
In November 2008, ADOT and NPS personnel removed 1.5 miles of ROW fencing on
each side of US 89 within Wupatki National Monument, including the T-posts. This
unprecedented large-scale fence removal is anticipated to have a greater impact on
promoting pronghorn highway crossings than the section modified at MP 444.1 to 444.2.
Also, ROW fencing along a 0.5-mile stretch of US 89 immediately north of Wupatki
National Monument was modified in 2009 by ADOT and Babbitt Ranches. Due to the
presence of livestock here, the fence was pulled back from the ROW >100 yards to allow
pronghorn the opportunity to cross fences and roadways individually. Unfortunately, the
second phase of the pronghorn telemetry was completed when these fence modifications
were made, limiting the tracking of animal response. Twelve pronghorn were captured in
late November 2008 and instrumented with GPS receiver collars, seven on the east side
and five on the west side of US 89 to assess potential pronghorn response to the removal
of ROW fence. The GPS collars from these animals will be recovered in December 2010
and movements in relation to the fence modification will be analyzed in a separate report.
21
3.4 TRAFFIC VOLUME AND PRONGHORN DISTRIBUTION
The research team had access to traffic volume data from a permanent ATR programmed
to record hourly traffic volumes. ADOT’s Data Management Section assisted the
research team in installing the ATR in March 2007 at a central location in the study area,
just north of Wupatki National Monument. As was done for elk by Gagnon et al. (2007b)
and white-tailed deer by Dodd et al. (2009), the research team combined traffic and GPS
data by assigning traffic volumes for the previous hour to each pronghorn GPS location
using ArcGIS® Version 9.1 (ESRI, Redlands, California, USA). This allowed the team to
correlate traffic volumes with pronghorn actions/movements during any given one-hour
time interval.
The research team examined how the proportion of pronghorn relocations at different
distances from the highway varied with traffic volume by calculating the proportion of
relocations in each 330-ft (0.0625 mi) distance band, out to a maximum of 3,300 ft (0.625
mi), similar to Gagnon et al. (2007b) for elk and Dodd et al. (2009) for white-tailed deer
except that previous analyses were limited to 2,000 ft. To avoid bias due to differences in
the number of relocations for individual pronghorn, the proportion of relocations
occurring in each distance band for each animal was used as the sample unit, rather than
total relocations. The team then calculated a mean proportion of pronghorn relocations
for all animals within each 330 ft-distance band at varying traffic volumes
(vehicles/hour): <100, 101−200, 201−300, 301−400, 401−500, and 501−600 (Gagnon et
al. 2007b). Pronghorn distribution and highway impact were compared to those for elk
(Gagnon et al. 2007b) and white-tailed deer (Dodd et al. 2009).
3.5 PRONGHORN-VEHICLE COLLISIONS
To track wildlife-vehicle collisions (WVC) involving pronghorn, the research team
primarily relied on accident report forms provided by Department of Public Safety (DPS)
highway patrolmen in the Flagstaff District, including the recording of roadkills where no
accident was reported. This information was augmented by periodic searches of the
highway corridor by the research team for evidence of WVC. WVC records were
compiled and summarized by highway reconstruction section by year. Lastly, ADOT’s
long-term statewide roadkill database (1990−2006) was queried for past WVC involving
pronghorn along the study stretch of US 89.
3.6 IDENTIFICATION OF PASSAGE STRUCTURE SITES
Of the 28 miles of US 89 (280 0.1-mile segments) within the study area, 12 miles (MP
430.0 to 442.0) were upgraded to a 4-lane divided highway in 1999. It is expected that
the remaining 16-mile length from MP 442.0 to 458.0 will be upgraded in the future.
FHWA guidelines permit ADOT to extend active construction activities up to 5% of the
project length in each direction beyond project limits. Therefore, with the 42 miles (MP
442.0 to 484.0) of reconstruction addressed under the final US 89 Antelope Hills –
Junction US 160 Environmental Assessment (ADOT 2006), construction activities could
extend 2 miles beyond the south end of the reconstruction zone, to MP 440.0. As such,
22
the research team assessed the potential for pronghorn passage structure sites from MP
440.0 to 458.0 and excluded the remainder of the study area to the south.
Sawyer and Rudd (2005) identified several important considerations for locating the most
suitable sites for pronghorn passage structures. In its assessment of potential passage
structure sites, the research team considered each criterion identified by Sawyer and
Rudd (2005), but recognized that the 0.1-mile segment scale was too small and
cumbersome to discern and analyze differences among segments. Dodd et al. (2006,
2007b) reported that the 0.6 mile (1 km) scale was optimum for making recommendations
for wildlife passage structures based on telemetry or WVC data. Making recommendations
at this scale also allows ADOT engineers latitude to determine the best technical location
for passage structures along the segment. Thus, for analysis of the criteria identified
by Sawyer and Rudd (2005), the team aggregated the 180 0.1-mile segments from
MP 440.0 (one 0.1-mile segment added) to 458.0 into 30 0.6-mile segments for analysis.
Sawyer and Rudd (2005) identified pronghorn abundance as a primary criterion for the
consideration of passage structure sites. The research team applied this observation on
the entire study stretch of US 89 and separately on individual segments. Sawyer and
Rudd (2005: 17) stressed that passage structures were more appropriate in linking
populations with “abundant numbers (i.e., hundreds)” and exhibit a high likelihood of
encountering passage structures, than small isolated populations that may not benefit to
the same degree. Since the pronghorn population adjacent to US 89 exceeds 600
animals, with the herds on both sides of the highway still viable and reproducing, the
research team determined that there is a sufficient population to evaluate passage
structure sites. Thus, the team used the other segment-specific criteria identified by
Sawyer and Rudd (2005) with minor modifications to rate each of the 30 0.6-mile
segments, considering GPS telemetry findings with other pertinent factors, as follows:
Pronghorn distribution − this rating was based on the mean (of 0.1-mile segments
within each 0.6-mile rating segment) number of different GPS-collared pronghorn
relocated within the 0.3-mile approach zone on either side of US 89. Ratings
were:
0 No animals approaching
1 1−2 animals approaching
2 3−5 animals approaching
3 6−8 animals approaching
4 9−10 animals approaching
5 >10 animals approaching
Pronghorn approaches – this criterion was considered the most important and
indicative of where animals potentially would approach and cross US 89 via a
passage structure, and was based on the mean number of approaches for the six
0.1-mile segments on both sides of the highway. Ratings were:
23
0 No approaches
1 1−10 approaches
2 11−20 approaches
3 21−30 approaches
4 31−40 approaches
5 41−60 approaches
6 >60 approaches
Land status – this criterion reflected the ability to conduct construction activities
outside the ADOT ROW, such as creating approaches with fill material for
overpasses. Ratings were:
0 State Trust
1 Private
2 Federal – NPS (natural ecosystem focus)
3 Federal – USFS (multiple-use focus)
Human activity – ideally, no human activity should occur within the vicinity of a
passage structure; however, road access, businesses, visitor pullouts, and other
activities do occur adjacent to US 89. Ratings were:
0 Significant human activity (business, housing, etc.)
1 Moderate human activity (access road, visitor pullout)
2 Limited human activity
3 No human activity
Fencing – fencing, especially of the highway ROW has a significant impact on
permeability. This criterion relates to the ability to eliminate or mitigate fencing
associated with a potential passage structure, and is closely tied to land
ownership. Ratings were:
0 Private lands near homes/businesses
1 State Trust land
2 Private land
3 Private lands with cooperative landowner (Babbitt Ranch)
4 USFS lands
5 NPS lands (no livestock grazing, no fencing needed)
Topography – the ability to situate overpasses oriented along existing ridgelines
that pronghorn can traverse, or locate underpasses in association with wide gentle
drainages is desirable. Ratings were:
0 Terrain not suited for a passage structure (steep, broken)
1 Topography marginal for a passage structure (flat)
2 Topography could accommodate a passage structure (drainage)
3 Topography ideally suited for passage structure (ridgeline or wide,
gentle drainage or basin)
24
Median width (not identified in Sawyer and Rudd 2005) – the selected alternative
median width has a large bearing on the potential distance that a passage structure
would need to span, as well as the distance animals would have to traverse in
crossing the highway. Ratings were:
1 84-ft median planned or existing
2 30-ft median planned
3 No median planned
25
4.0 RESULTS
The research team instrumented and tracked 37 pronghorn (20 females, 17 males) with
GPS receiver collars from January 2007 to December 2008. Most animals were captured
in January and December 2007 as part of two separate telemetry phases, though two
animals were captured in August 2007 following the independent death of two females
(both apparently by dogs near private land). The team was able to instrument 19
pronghorn with collars on the west side of US 89 and 18 on the east side, meeting its
objective for distributing collars equally.
GPS collars were affixed to pronghorn an average of 266.2 days (±30.6 SE), during
which time the collars accrued 118,181 GPS fixes (Figure 6) for a mean of 3,194.1
fixes/pronghorn (±376.1). Of the GPS fixes, a mean of 86.6% were 3-D fixes and 13.4%
were lower accuracy 2-D fixes.
4.1 PRONGHORN MOVEMENTS, DISTRIBUTION, AND APPROACHES
4.1.1 Pronghorn Movements and Distribution
Of the GPS fixes accrued (Figure 6), 1,125 (1.0%) occurred within 0.15 mile of US 89, or
an average of 33.1 (±7.9) fixes/animal; eight pronghorn did not approach the highway to
within 0.15 mile. Within 0.30 mile of the highway, there were 3,794 pronghorn fixes
(3.2% of all fixes) for an average of 102.5 (±20.4) fixes/animal; six animals did not
approach to within 0.30 mile of US 89. Pronghorn approached to within 0.60 mile of the
highway 10,230 times (8.7% of all fixes), with a mean of 276.5 (±46.6) fixes/animal; four
pronghorn never approached to within 0.60 mile of US 89 during the study.
Over the duration of GPS tracking, pronghorn (n = 37) travelled an average of 3.2 miles
(±0.07) each day. Males (n = 17) travelled slightly further each day (3.3 miles; ±0.10)
than females (3.1 miles; ±0.10; n = 20), though the difference was not significant (P =
0.177).
MCP home ranges for all pronghorn with sufficient fixes to estimate home ranges(n = 30)
averaged 71.1 mile2 (±6.8). Male (n = 12) MCP home ranges (76.7 mile2; ±10.3) were
considerably large than females’ (n = 18) which averaged 66.1 mile2 (±9.0); however, the
difference was not significant (P = 0.668). There was no significant difference between
home ranges on each side of US 89 (P = 0.605).
4.1.2 Pronghorn Highway Crossings and Permeability
Only one GPS-collared pronghorn crossed US 89 during the nearly two years of tracking,
a female that crossed 12 times in June 2007; none of the other 36 collared pronghorn
crossed the highway. This animal was apparently travelling to water tanks on the east
side of the highway from the west side of US 89 where it resided the vast majority of the
26
Figure 6. Distribution of GPS fixes for 37 pronghorn accrued from 2007 to 2008
adjacent to US 89, Arizona. Each color represents an individual collared pronghorn.
time. The crossing rate for this animal was 0.05 crossings/day, and the crossing rate
averaged 0.001 crossings/day among 30 pronghorn that approached US 89 to within 0.15
mile. The mean pronghorn passage rate was negligible (0.006 crossings/approach; n =
30).
4.1.3 Pronghorn Approaches
The frequency of approaches that pronghorn made to within 0.30 mile of US 89 yielded
considerably more information than crossings to assist in determining the locations of
potential passage structures. The number of approaches differed considerably on each
side of the highway, though not significantly (P = 0.103). There was three times the
number of approaches from the west side compared to the east side. On the west side, 18
27
pronghorn approached the highway 2,875 times (Figure 7), for a mean of 159.7 (±34.3)
approaches/animal. There was an average of 9.0 approaches/0.1-mile segment, and a
range of 0 to 134 approaches among the segments. Among segments, the number of
different approaching pronghorn ranged from 0 to 13 and averaged 1.7. The observed
approach distribution did not occur in a random distribution (Kolmogorov-Smirnov d =
0.309, P< 0.001).
On the east side of US 89, 13 pronghorn approached the highway 952 times (Figure 7),
with a mean of 73.9 (±21.5) approaches/pronghorn. Among 0.1-mile segments,
pronghorn approaches averaged 3.1/segment and ranged from 0 to 55 approaches. The
number of different pronghorn approaching the highway at a given segment from the east
side ranged from 0 to 5, and averaged 0.8/segment. The observed distribution of
approaches from the east also differed from a discrete random distribution (Kolmogorov-
Smirnov d = 0.249, P< 0.001).
On the west side of the highway, the SDI-weighted pronghorn approaches totaled
3,435.2, and averaged 11.1/segment (Figure 8). The weighted distribution of approaches
(Figure 8) lacked several of the peaks in approach frequency for unweighted approaches
between segments 89−111 and 199−221 (Figure 7), as relatively few animals accounted
for these unweighted approach peaks. Conversely, the large peak in approaches between
segments 122 and 155 (Figure 7) increased substantially when weighted by SDI (Figure
8) reflecting the large number of different collared pronghorn that approached in this
area, 15 of the total 18 that approached on the west side of US 89. Beyond segment 220
(through 310) there were no weighted approaches (Figure 8).
On the east side of the highway, SDI-weighted approaches decreased from the
unweighted approach total to 682.2, averaging 2.2 approaches/segment (Figure 8), and
reflected the fact that many of the approaches were made by relatively few animals. This
was particularly true for the large peak in approaches between 50 and 57 (Figure 7). Like
the west side of the highway, there were no weighted approaches beyond segment 220.
With weighted approaches by pronghorn from both sides of US 89 combined, totaling
5,035.2 approaches (16.2/segment), a significant peak accounting for nearly half (47%)
of the approaches occurred at the north end of the Coconino NF between segments 130
and 150 (MP 441-442), or only 7% of the length of the study area (Figure 9). Of 31 total
pronghorn that approached the highway, 22 (71%) approached US 89 in this 2-mile zone.
4.1.4 Linear Approach Distance along Highway
GPS-collared pronghorn (n = 31) were recorded within 0.30 mile of the highway along a
mean linear distance of 6.5 miles (±0.8) adjacent to US 89. The linear distance in which
pronghorn approached the highway on the east side (7.9 mile; ±1.2; n = 13) was greater
than animals approaching on the west side (5.6 mile; ±1.1; n = 18), though the difference
was not significant (P = 0.169). Likewise, though the mean linear distance in which
approaches occurred adjacent to the highway by males (7.7 miles; ±1.5; n = 13) was
higher than by females (5.7 miles; ±0.9; n = 18), the difference was not significant (P =
28
Figure 7. Frequency distribution among 0.1-mile segments of approaches to within 0.3
mile of US 89 made by 18 pronghorn on the west side of the highway (top) and by 13
pronghorn on the east side of the highway (bottom). Both distributions differed from a
discrete random distribution.
29
Figure 8. Frequency distribution among 0.1-mi segments of Shannon diversity index-corrected
weighted approaches to within 0.3 mi of U.S. Highway 89, Arizona made by 18 pronghorn on
the west side of the highway (top) and by 13 pronghorn on the east side of the highway (bottom).
Weighted approaches reflect animal approach frequency, number of approaching animals, and
equity in distribution among approaching pronghorn.
30
= 0.231). This linear distance reflects the extent that pronghorn approach and use the
habitat adjacent to US 89, and has a bearing on the number and spacing of potential
passage structures.
4.2 PRONGHORN MOVEMENTS AND FENCING REMOVAL
At and adjacent to the 0.1-mile segment (172; MP 444.1 to MP 444.2) where barbed-wire
fencing was removed from ROW fence T-posts within Wupatki National Monument, no
pronghorn crossings were recorded. At the segment where fence was modified and 0.1
mile on either side, approaches to within 0.3 mile of the highway averaged 6.3
approaches/segment. This compared to a mean of 7.0 approaches/segment in the adjacent
half-mile stretch to the north of the modified fence, and a mean of 5.8
approaches/segment a half mile to the south. Similarly, there was no discernable increase
in approaches to within 0.15 mile of US 89 that could better reflect an “intent” or attempt
to cross the highway; two approaches occurred at the 0.1- mile segment where fence was
modified compared to a mean of 2.0 approaches/segment a half mile to the south, and 1.6
approaches/segment for a half mile to the north. As such, it does not appear that
pronghorn responded with increased highway crossings or approaches to this limited
section where fencing was modified. This underscores the importance of the effort to
remove fence along three miles of ROW in November 2008 and its prospect for
achieving improved pronghorn passage, as well as the two-year commitment to monitor
pronghorn response to this fence removal.
4.3 TRAFFIC VOLUME AND PRONGHORN DISTRIBUTION
The distribution analysis was based on 10,100 pronghorn GPS relocations recorded
within 3,300 ft of US 89. Frequency distributions of mean probabilities of pronghorn
occurring in distance bands showed minimal shift in distribution away from the highway
at increasing traffic volume until traffic reached 500 vehicles/hr (Figure 10). Among all
traffic volumes up to 500 vehicles/hr, mean distribution probabilities were constant
(Table 1), varying minimally (<15%) among traffic volume classes. At distances
between 0 to 990 ft from the highway, the mean distribution probabilities at traffic
volumes up to 500 vehicles/hr were <0.15. At traffic volumes between 500 and 600
vehicles/hr, mean pronghorn distribution probabilities between 1,320 and 2,970 ft
declined 17% compared to lower traffic volume classes, and increased 73% in the 2,970
to 3,330 ft class (Table 1, Figure 10). The mean probability of female pronghorn
occurring within 1,980 ft of US 89 across all traffic volumes (0.45) was slightly lower
than the probability for males (0.49). The most apparent difference between sexes
occurred at 3,300 ft from the highway, where the female distribution probability across
all traffic volumes (0.17) was higher than the mean male probability (0.11).
4.4 PRONGHORN-VEHICLE COLLISIONS
During the project from 2007 to 2008, no WVC involving pronghorn were recorded by
DPS highway patrolmen or the research team along US 89. Further, no pronghorn
Figure 9. Combined frequency distribution among 0.1-mile segments of weighted approaches to within 0.3 mile of US 89 made by 31
pronghorn on both sides of the highway.
31
32
records were found in ADOT’s long-term statewide roadkill database for the period
1990−2006.
4.5 IDENTIFICATION OF PASSAGE STRUCTURE SITES
The distribution of weighted approaches by pronghorn along US 89 (Figure 9) alone
suggests locations where passage structures might be appropriate. When combined with
other criteria such as identified by Sawyer and Rudd (2005), ratings of 0.6-mile segments
between MP 440.0 to 458.0 for suitability to passage structures ranged from 5−26 points
(Appendix A, Figure 11). The highest rated 0.6-mile segment (0.1-mile segments 142-
147) on the Coconino NF boundary south of Antelope Hills corresponded to the stretch of
highway with the highest mean weighted pronghorn approaches (116.1
approaches/segment), highest mean number of different pronghorn (10.3/segment), and
favorable land ownership, all of which make this site highly suited for a pronghorn
passage structure. The second highest rated segment was immediately to the north (148-
153), scoring 24 points. The adjacent segment corresponded to the area encompassing
the private lands at Antelope Hills with relatively few weighted approaches
(23.2/segment) and different pronghorn (3.8/segment), coupled with high human activity
and poor prospect for addressing fencing needs, all contributing to its low rating for
passage structure suitability (Figure 11; Appendix A). The next four highest rated
segments (18-19 points) were those on Wupatki National Monument (segments 160-183)
up to the segment with the main park entrance road and visitor pullout that rated lower
(17 points). North of Wupatki National Monument between segments 196−219, three of
four 0.6-mile segments rated >10 points (Figure 11), with the one at segments 202−207
corresponding to the minor peak in pronghorn approaches (Figure 9). Beyond segment
220, there were no weighted crossings and no more than a single different pronghorn that
approached the 0.1-mile segments.
Table 1. Mean combined probabilities that GPS-collared pronghorn (n = 31) found
within distance bands from the highway at varying traffic volumes. Probabilities were
determined from pronghorn telemetry and traffic counting conducted along U.S.
Highway 89, Arizona, from 2007−2008.
Distance Probability of occurring in distance band by traffic volume (vehicles/hr)
from
highway (ft) <100 100−200 200−300 300−400 400−500 500−600
300−990 0.15 0.13 0.15 0.14 .014 0.14
990−1,980 0.33 0.35 0.35 0.33 0.33 0.27
1,980−2,970 0.40 0.40 0.38 0.38 0.37 0.34
2,970−3,300 0.12 0.13 0.11 0.15 0.16 0.26
33
Figure 10. Mean probabilities that GPS-collared pronghorn (n = 31) occurred within
each 330-ft distance band from the highway at varying traffic volumes: a) <100, b)
100−200, c) 200−300, d) 300−400, e) 400−500, f) 500−600 vehicles/hr.
34
Figure 11. Ratings of suitability for pronghorn passage structures based on seven criteria
by 0.6-mile segment between US 89 mileposts 440.0 and 458.0 (Appendix A). Red bars
denote the three recommended segments for pronghorn passage structures based on
ratings, weighted approaches, and approximate 3.2-mile spacing between passage
structures. Gray bars denote segments where no weighted pronghorn crossings were
recorded.
35
5.0 DISCUSSION
Dodd et al. (2006) advocated utilizing WVC and roadkill data where it exists as a
surrogate to costly GPS telemetry movement information to plan and identify locations
for wildlife passage structures; they found that the spatial incidence of WVC was
strongly associated with GPS-determined highway crossings. However, in the instance
of a species like pronghorn where highways constitute passage barriers to the degree that
pronghorn-vehicle collisions do not occur, GPS telemetry data is essential to developing
informed, data driven recommendations for passage structure placement. And where
traditional VHF-telemetry studies were instrumental in first demonstrating the degree to
which northern Arizona highways constituted barriers to pronghorn movement
(Ockenfels et al. 1997, van Riper and Ockenfels 1998, Bright and van Riper 2000, Hart et
al. 2008), these studies provided only limited insights on the best locations for potential
passage structures when compared to GPS telemetry. Dodd et al. (2007d) stressed how
GPS telemetry has revolutionized wildlife movement studies, particularly those intended
to quantify wildlife permeability across highways, as GPS yields tremendous amounts of
unbiased movement data. In this study, the mean number of GPS relocations for
individual pronghorn (n = 3,194) exceeded the relocations of all animals combined on
most previous Arizona VHF telemetry projects, with the exception of Ockenfels et al.
(1994) that relocated 47 animals 4,996 times.
5.1 PRONGHORN PERMEABILITY
With the insights gained from previous northern Arizona telemetry studies (Ockenfels et
al. 1997, van Riper and Ockenfels 1998, Bright and van Riper 2000, Hart et al. 2008), the
research team expected that the US 89 pronghorn passage rate would be low compared to
elk (Dodd et al. 2007d, 2009; 0.43−0.88 crossings/approach) and white-tailed deer (Dodd
et al. 2009; 0.03−0.16 crossings/approach). However, the negligible pronghorn passage
rate (0.006 crossings/approach) where only one of 30 animals that approached US 89
ultimately crossed was even lower than anticipated. The lone female crossed at the driest
time of the year (June 2007), apparently crossing the highway six times to utilize stock
tanks on the east side and returned each time. Unfortunately, these movements occurred
outside of the breeding season such that no genetic benefit was realized. US 89 can be
considered a near-total barrier to the passage of pronghorn, and thus has effectively
subdivided the GMU 7 herd into two isolated populations. The future reconstruction of
US 89 between MP 442.0 and 458.0 to a 4-lane divided highway with an 84-ft median
will certainly exacerbate this barrier effect, as predicted by wildlife highway avoidance
models (Jaeger et al. 2005, Hart et al. 2008).
The impact of isolation on the separate herds on each side of US 89 is difficult to assess.
Current genetic assessment of the populations on each side of US 89 and comparison to
other populations is now underway. This assessment is similar to those done elsewhere
for bighorn sheep (Epps et al. 2005) and meso-carnivores (Riley et al. 2006), and will
help quantify the impact of highway-induced genetic isolation and drift. Genetic samples
collected during the capture of pronghorn on this project are being analyzed as part of a
cooperative effort between ADOT, AGFD, and Northern Arizona University’s
36
Department of Biology (ADOT Project SPR-659, Genetic Variation of Pronghorn Across
US Highway 89 and State Route 64). This genetic assessment will constitute a baseline
from which to conduct future assessments of the genetic effectiveness of passage
structures, argued for by Corlatti et al. (2009) as being warranted to justify the high cost
of such structures.
The GMU 7 pronghorn population may be exhibiting symptoms of isolation. Though the
populations on each side of US 89 are largely isolated from each other, there likely
remains some level of gene flow and interaction with pronghorn further to the west and
east, though the distance that GPS-collared animals travelled away from the US 89
corridor over two years was limited (Figure 6). The research team’s observations, as well
as those of other pronghorn biologists familiar with the US 89 populations (R. Ockenfels
and B. Holton, personal communications) were that the eastern herd has noticeably
declined over the past ten or more years. During the initial capture effort in January
2007, a winter snowstorm pushed and concentrated over 350 animals from the northern
slopes of the San Francisco Peaks to within a mile of the west side of US 89. On the east
side, only 30-40 different animals were found within five miles of US 89, similar to the
number found in December 2007 during the subsequent second capture. When animals
were captured for previous VHF telemetry studies in 1992−1994 (Ockenfels et al. 1997,
van Riper and Ockenfels 1998), “hundreds” of pronghorn were seen on the east side and
within five miles of US 89 (R. Ockenfels, personal communication). The pronghorn
recruitment rate (fawns:doe) on the east side of US 89 has averaged 40% lower than on
the west side since 2003. This is a source of concern for a population exhibiting a
downward population trend (AGFD unpublished Pronghorn Hunt Recommendations,
Game Branch, Phoenix) and not benefitting from population replenishment and genetic
flow from the potential “source” population on the west side of US 89.
In spite of these population and isolation concerns, the findings of this study serve to
illustrate that the population’s numbers adjacent to US 89 remain sufficient to benefit
from potential passage structures that would promote connectivity, an important
consideration stressed by Sawyer and Rudd (2005). Furthermore, the movement,
distribution, and weighted approach data suggest that it is likely that pronghorn would
encounter and use one or more passage structures, which would ultimately successfully
enhance permeability.
5.2 TRAFFIC VOLUME AND PRONGHORN DISTRIBUTION
Pronghorn distribution among all distances and across all traffic volumes up to 500
vehicles/hr remained constant. Only at 500−600 vehicles/hr (equivalent to
12,000−14,400 AADT) did a change in distribution occur. At this traffic volume the
probability of pronghorn being 3,300 ft from US 89 was nearly double that at 100
vehicles/hr or lower. Mean probabilities of occurrence of white-tailed deer also showed
minimal shift in distribution away from SR 260 at increasing traffic volume, with the
mean probability of a deer occurring within 660 ft of the highway remaining constant
from approximately 0.32 at <100 vehicles/hr to 0.28 when traffic was >600 vehicles/hr
(Dodd et al. 2009). In contrast to both pronghorn and deer, elk along SR 260 exhibited
37
dramatic shifts in distribution across distance bands with increasing traffic, including a
>50% reduction in the probability of being within 660 ft of the roadway as traffic
increased from <100 to 600 vehicles/hr. However elk exhibited higher probabilities of
occurrence closer to the road when traffic volumes were at their lowest (Gagnon et al.
2007b). Elk utilized lush meadows adjacent to SR 260 for feeding (Manzo 2006; Dodd et
al. 2007b) and were relocated within 0.6 mile of the highway at twice the expected
frequency of occurrence (Dodd et al. 2007d). In the absence of such attractive habitats
adjacent to US 89, pronghorn appear to lack an incentive or attractant to “tolerate” even
the impact of relatively low traffic volumes. Whereas Reeve (1984) reported that regular
vehicular traffic produced minimal disturbance among pronghorn due to habituation, the
research team believes that pronghorn along US 89 are consistently negatively impacted
by traffic volume even at low levels, which is seldom during the daytime when
pronghorn are most active (Figure 3). Daytime traffic volumes along US 89 typically
exceed the 10,000 vehicles/day level where highways become strong barriers to wildlife
passage and traffic repels animals from the roadway, as hypothesized by Mueller and
Berthoud (1997).
Just as Reeve (1984) found females with young to be more sensitive to vehicular traffic
and Gavin and Komers (2006) reported that female pronghorn in spring and herds with
young exhibited higher proportions of vigilant behavior, a 54% higher proportion of
female relocations compared to males occurred at the furthest distance (3,300 ft) from US
89 that was assessed. This suggests that females may be more sensitive to traffic-associated
impacts from the highway.
5.3 STRATEGIES TO PROMOTE PRONGHORN PERMEABILITY
As detailed by Sawyer and Rudd (2005), several factors support the continued pursuit and
development of strategies for wildlife passage structures along the US 89 study area,
including:
• Pronghorn populations on each side of US 89 are sufficient in numbers
(particularly on the west side) and remain viable with reproduction, though they
may be exhibiting signs of isolation indicating the future action to promote
permeability is warranted; pronghorn population status warrants action and it is
expected that the population will benefit and respond to management actions.
• US 89 represents a near-total barrier to the passage of pronghorn across the
highway; without intervention, especially as the highway is upgraded, the impact
of the highway barrier effect will increase and further limit pronghorn movement.
• The distribution and movement of pronghorns along US 89, the numbers for
weighted approaches, and the counts for different animals using highway
segments suggest that pronghorn will likely use passage structures. Peak use
zones and areas of pronghorn concentration near US 89 ensure that a significant
portion of the population will encounter passage structures once built.
38
• Land ownership is conducive to potential passage structure construction and
integrated management of ROW fences, both of which are critical to the success
of promoting permeability.
In developing its strategies to promote pronghorn permeability along US 89, the research
team addressed the following considerations:
• Number of passage structures and spacing needed to accommodate pronghorn
passage.
• Locations and priorities for potential passage structures.
• Role of ROW fencing and options.
• Types of passage structures and specific design criteria.
5.3.1 Number and spacing of passage structures
The spacing of wildlife passage structures has a potentially significant impact on their
ability to promote highway permeability (Olsson 2007, Bissonette and Adair 2008, Dodd
et al. 2009). Bissonette and Adair (2008) recommended spacing of 2.0 miles between
passage structures to accommodate pronghorn permeability based on isometric scaling of
home ranges. Using this criterion for the stretch of US 89 evaluated for passage
structures (MP 440.0 to 458.0), approximately nine passage structures would be required.
If the target highway stretch was shortened by 9 miles to eliminate the 0.1-mile segments
where no weighted approaches by pronghorn occurred (segments 221−310), four to five
passage structures would be required to promote pronghorn permeability.
The mean linear distance that pronghorn travelled adjacent to US 89 over the duration of
the project was 6.5 miles, while the daily distance traveled by individual pronghorn
averaged 3.2 miles. These movements reflect the “flattening” effect that US 89 has on
the configuration of pronghorn home ranges and the linear distance travelled along this
impermeable highway corridor (Figure 12). Ockenfels et al. (1997) reported elongated
home ranges reflecting constriction along transportation corridors compared to more
typical pronghorn home range shapes (O’Gara and Yoakum 1992). Clevenger et al.
(2001) noted that animals travelled parallel to a highway after encountering the roadway
corridor. On the other hand, recommendations made by Bissonette and Adair (2008)
reflect uniformly shaped home ranges that were bisected by highways.
The research team believes that 3.2-mile spacing between passage structures more
realistically reflects empirical pronghorn movements along US 89. This recommended
spacing reflects the mean daily distance traveled by pronghorn and is half the mean linear
distance travelled along US 89 by collared pronghorn. This distance ensures that a
passage structure would be encountered no further than 3.2 miles in either direction along
the average linear travel distance. Using a passage structure spacing of 3.2 miles over the
entire 16.8-mile stretch of highway, five passage structures would be required.
39
Excluding the 9-mile stretch of highway where no weighted pronghorn approaches
occurred, two to three passage structures would be required to promote pronghorn
permeability.
Figure 12. Comparison of GPS fix distributions for three representative pronghorn
adjacent to US 89, with the pronghorn on the east side of the highway (yellow) never
approaching the highway and thus not flattening the distribution linearly along the
impermeable barrier as with the two pronghorn on the west side of the highway.
40
5.3.2 Locations and Priorities for Potential Passage Structures
Applying a spacing distance of 3.2 miles, priority sites for passage structures were
identifed based on the rating of 0.6-mile segments. In identifying locations for potential
passage structures, the research team excluded the 9-mile northernmost stretch of US 89
where no weighted pronghorn approaches occurred (Figure 7).
Coconino National Forest - Antelope Hills Site
Both the distribution of weighted pronghorn approaches (Figure 9) and ratings by 0.6-
mile segment (Figure 11; Appendix A) unequivically show segments 136-141 (MP
440.6−441.1) to be the best stretch for a passage structure along US 89. This segment
exhibited the largest peak in approaches and numbers of different pronghorn approaching
the highway (Figure 9; Appendix A). This site appears to function as a “crossroads” for
pronghorn moving adjacent to US 89 in either direction; 13 of 18 (72%) GPS-collared
animals on the west side had approaches within this 0.6-mile segment, and six of 13
(46%) on the east side.
The concentration of pronghorn at this site likely reflects several factors. First, the area is
situated on the ecotone or transition between juniper woodland and short grass
prairie/grasslands (Figure 13). This zone reflects a transition from the fairly broken
terrain to the south to the generally flat terrain vegetated by grasslands to the north
(Figures 9 and 13). Though pronghorn are adapted to using open flat to undulating or
rolling topography (Yoakum 1980), they do use broken and steeper terrain (Ockenfels et
al. 1994). This vegetative and terrain transition zone near the Coconino NF boundary
affords pronghorn refuge from winter storms and winds amongst junipers and behind
leeward ridges, cinder cones, and other terrain. During summer, woodland cover is used
for shading (Ockenfels et al. 1994), though pronghorn typically avoid dense woodlands
(Ockenfels et al. 1994, Bright and van Riper 2000). This transition area also reflects
higher vegetative diversity and composition, which has a strong influence on pronghorn
populations (Ockenfels et al. 1994, Yoakum 1980). Areas with high shrub diversity have
been shown to be important for fawning cover (Bright and van Riper 2000). The
concentration of pronghorn in the vicinity of the Coconino NF boundary could also
reflect differences in livestock management practices and potential competition with
pronghorn on USFS, State Trust, and private lands (McNay and O’Gara 1982; B. Holton,
unpublished report, NPS, Flagstaff, AZ).
This area of highest concentrated pronghorn distribution is along the US 89 section that
was reconstructed to a 4-lane divided highway in 1999 (Figure 13). As such, erecting a
passage structure here would constitute a “retrofitting” effort, though this would be
permissible if done as part of an adjacent highway reconstruction project. The terrain at
this site is generally a large west to east running ridge that slopes downward to the east;
there are large cut slopes on the west side of the highway (Figure 13). This site would be
conducive to the construction of an overpass for pronghorn passage.
41
Wupatki National Monument Site
Approximately three miles to the north of the Coconino National Forest – Antelope Hills
site is the 0.6-mile segment (segments 172−177; MP 444.2−444.7) situated amongst the
four relatively high-rated segments on Wupatki National Monument (Figure 11). Though
this segment lacks the high frequency of weighted pronghorn approaches that occurred to
the south, there were nonetheless approaches made by 11 different pronghorn. One of
this site’s most attractive features is the ease of addressing ROW fencing issues since no
livestock graze here; the majority of the ROW fence on Wupatki National Monument has
already been removed to facilitate pronghorn passage. The high quality of pronghorn
habitat associated with the location also makes this site a favorable location for a crossing
structure.
Roadway reconstruction plans for this area include a 4-lane divided highway with a 30-ft
median, 54 ft less than the median planned north of Wupatki National Monument (ADOT
2006); this is another attractive aspect of this site as it requires a shorter span for a
potential passage structure. The terrain within this segment ranges from a gentle shallow
basin up to a gentle incline along the highway that culminates at a fairly flat ridgeline
running west to east, with cut slopes on each side of the highway (Figure 14). This
segment has two potential passage structure sites, one in the basin near a drainage that
would support a wide underpass/viaduct type structure and the other at the ridge crest that
would be more conducive to an overpass (Figure 14).
Babbitt Ranch Site
A peak in weighted pronghorn approaches occurred at segments 202−207 (MP
447.2−447.7; Figure 9), at which five different GPS-collared pronghorn approached the
highway. This segment spans both private lands owned by Babbitt Ranches and State
Trust land. The terrain in the segment is a broad flat basin typical of open pronghorn
habitat (Figure 15). The segment lacks the topographic relief found at the other two
priority passage structure sites (Figure 15). The rating for this segment (Figure 11)
reflects the fact that Babbitt Ranches has taken a highly proactive role in pronghorn
management and has expressed a desire to cooperate in the evaluation of potential
passage structures and further modification of ROW fencing. A viaduct-type structure
would be appropriate at this relatively flat basin site.
5.3.3 Role of ROW Fencing and Options
Hart et al. (2008) concluded that the lack of pronghorn response to fence modification
treatments intended to promote passage across the railroad corridor through Petrified
Forest National Park reflected several factors, including heavy rail traffic that deterred
animals from approaching the corridor. However, the relatively short extent of the
modified fence (0.6 mile) and the fact that T-posts were not removed contributed to the
lack of measurable response by pronghorn. Thus, it is not surprising that pronghorn did
not cross or approach US 89 at the 0.1-mile section where the barbed wire was removed
but T-posts were left standing, though this fence had been modified for several years
compared to the fence modifications Hart et al. (2008) assessed that were in place for only
six months. The removal or modification of nearly two miles of ROW fence on Wupatki
42
Figure 13. Aerial view (top) and enlarged oblique view (bottom) from GoogleEarth©
depicting the proposed Coconino NF - Antelope Hills pronghorn passage structure site on
US 89 between MP 440.6 and 441.1.
43
Figure 14. Aerial view (top) and enlarged oblique view (bottom) from GoogleEarth©
depicting the potential Wupatki National Monument pronghorn passage structure sites
between US 89 MP 444.2 and 444.7.
44
Figure 15. Aerial view (top) and enlarged oblique view (bottom) from GoogleEarth©
depicting the Babbitt Ranch site between US 89 MP 447.2 and 447.7.
45
National Monument and Babbitt Ranches should provide a more thorough and conclusive
evaluation of the ability to promote highway passage by eliminating or reducing the
barrier effect of fences.
Whereas pronghorn evolved in open plains/grassland environments where speed and
mobility was their defense against predators (Hart et al. 2008), this species has exhibited
limited ability to adapt to fences like elk, deer, and other species have. And whereas
fencing has been instrumental in preventing at-grade highway crossings and funneling
animals to passage structures to reduce WVC and promote permeability (Clevenger et al.
2001, Dodd et al. 2007c), such an approach may not be necessary for pronghorn since
they make few at-grade highway crossings and have few collisions with vehicles in
Arizona. Sawyer and Rudd (2005:18) stressed the advantage of avoiding fences
altogether in association with passage structures to promote pronghorn use. They stated
that “ideally, a crossing structure should be located in an area with no fencing. If fencing
is required, then (the) crossing structure(s) should be located in area(s) where fence
design is pronghorn-friendly … and do not inhibit pronghorn movements to and from the
structure.” With the absence of livestock on Wupatki National Monument, ROW fencing
is not needed, a significant advantage for promoting pronghorn permeability. On the
Coconino NF and Babbitt Ranch lands, creative approaches such as pulling fences back
0.25 to 0.5 mile or resting pastures and removing fencing (B. Cordasco, Babbitt Ranches,
personal communication) or raising/removing the bottom strand of barbed-wire fences
(Hart et al. 2008) could minimize the impact of ROW fences. ROW fencing in
association with passage structures is not needed to preclude at-grade pronghorn
crossings of US 89. Fencing would play less of a physical funneling role (e.g., compared
to elk; Dodd et al. 2007c) than providing a visual clue as to a path of least resistance
across the highway barrier, provided no fencing is used at the mouth of the passages.
5.3.4 Types of Passage Structures and Specific Design Criteria
The focus of this project was to first identify the need for passage structures along US 89,
which the research team believes is apparent (e.g., Figure 9), and then determine best
locations for passage structures. Determining the appropriate types of structures for
application at these sites is best left to ADOT’s engineers. This is partly the reason for
evaluating 0.6-mile segments: they provide technical latitude in specific site placement.
And though the specific site characteristics will dictate what type of potential structure
(e.g., underpass, viaduct, overpass) might be appropriate from engineering and cost
standpoints, a few general criteria are important for consideration as specific structures
are considered.
Structural design characteristics have a significant bearing on the eventual use and
acceptance of passage structures by wildlife (Foster and Humphrey 1995; Clevenger and
Waltho 2003; Gordon and Anderson 2003; Ng et al. 2004; Dodd et al. 2007b, 2009).
Most important is the requirement that any type of structure that is considered for
pronghorn passage be as open and wide as possible (Ruediger 2002, Sawyer and Rudd
2005), with special attention paid to avoiding obstructed line-of-sight views through or
across the structures (Foster and Humphrey 1995, Sawyer and Rudd 2005, Dodd et al.
46
2007a, 2009). This species’ adaptation to an open plains/grassland environment has
resulted in a strong survival reliance on visual stimuli and avoidance of dense habitats
and situations that restrict their view or mobility (Hart et al. 2008). While Yoakum
(2004) questioned the ability to achieve pronghorn use of passage structures across high
traffic volume roadways due to behavioral characteristics (e.g., highway avoidance),
Sawyer and Rudd (2005) concluded that properly designed and located structures could
be effective. They favored wide (>60 ft between bridge supports) and high (>24 ft) open-span
bridge/underpass structures to overpasses, and in recognizing the lack of insights for
pronghorn passage, believed underpasses to have wider application and lower cost, while
also helping address drainage needs. The research team stresses that topography and the
maximization of visual continuity for pronghorn are also critical concerns that may make
overpasses attractive and/or applicable along certain US 89 locales. Most wildlife
underpasses implemented along SR 260 that have proven so successful in promoting elk
and deer passage (Dodd et al. 2009) would not function well for pronghorn passage; nor
are similar topographic features prevalent along US 89 in which to situate underpasses, a
strong selling point made by Sawyer and Rudd (2005) in recommending underpasses.
Given a paucity of topographic situations in which to construct SR 260-type underpasses,
including at all three priority US 89 passage structure sites, the research team envisions
the potential application of either overpasses or open, elevated roadways/viaducts under
which animals pass (Figure 16). Aside from their greater cost, the main drawback of
such an underpass or viaduct would be traffic-associated noise emanating from the
elevated roadway.
With the impact evident from the high daytime traffic on US 89, including the visual and
noise impact on pronghorn (Mueller and Berthoud 1997), comprehensive measures to
reduce the traffic-associated impact could create “quiet zones” along the highway in the
areas of passage structures. Such quiet zones could facilitate pronghorn approaching
(and successfully crossing) the highway and play a potentially significant role in
promoting passage. A comprehensive set of measures should include incorporation of
highway design, noise barriers that don’t restrict movements, and pavement treatments
(Kaseloo and Tyson 2004). In conjunction with passage structure construction, for
instance, highway approaches to the structure could be recessed below grade to reduce
the noise impact while supporting overpass construction. Soil berms or sound walls
adjacent to passage structures may be warranted to help reduce traffic impact, as well as
to shield traffic from pronghorn sight and vice versa. Such barriers could reduce traffic
noise by as much as half, depending on their height (FHWA 2001). Further, shielding
vegetation atop berms could further reduce traffic-associated noise, as would rubberized
asphalt pavement near passage structures. Without a comprehensive effort to reduce
traffic’s noise and visual impact, the success of passage structures could be compromised
by continued deterrence of pronghorn from the highway at high traffic volumes (Mueller
and Berthoud 1997, Yoakum 2004).
One type of structure referenced by Sawyer and Rudd (2005) as having potential for use
as a pronghorn passage is CON/SPAN® pre-cast concrete arches, which can span 60 ft
47
Figure 16. Renderings of potential pronghorn passage structures that emphasize
openness and unobstructed views for crossing pronghorn, including an overpass
capitalizing on existing terrain (top) and an elevated roadway/viaduct over gentle terrain
(bottom).
48
and have various heights up to 24 ft and widths up to and exceeding 100 ft; additional
height can be achieved by extending the pedestal walls upon which the barrels rest.
Such a structure is potentially attractive for the Coconino NF- Antelope Hills site as it
can quickly be dropped into place in a retrofit application with minimal disruption to
traffic flow, is more cost effective than traditional bridge applications, and can maintain
existing sloping ridgeline integrity by side-by-side installation with arches of different
heights (Figure 17). Furthermore, the headwalls could be flared to maximize openness,
with a width of 100 ft or more to create a wide, open overpass on which pronghorn would
cross over US 89 (Figure 17).
A construction products company provided an estimate of the cost to erect a typical pre-cast
concrete arch overpass at the Coconino NF-Antelope Hills site (MP 440.9) and
conceptual plan sheets (Appendix B). The estimated cost for the construction of a 100-ft
wide overpass is estimated at $542,725 (≅ $700,000 including installation but excluding
fill atop the structure). This structure could be erected at the site in a matter of days.
Figure 17. Rendering of a CON/SPAN® pre-cast concrete arch application for a wildlife
overpass at MP 440.9 along US 89. The bottom rendering depicts separate 32-ft spans
(21 ft high) over each set of lanes, looking south, and the inset depicts the top of the
overpass, looking east (note: fencing would need to be erected atop the walls).
49
6.0 CONCLUSIONS AND RECOMMENDATIONS
This project used a data-driven approach to quantifying pronghorn permeability across
US 89, as well as determining the best locations for potential passage structures to
enhance permeability. The study was particularly important given the lack of WVC data
involving pronghorn, data that typically can serve as a surrogate to GPS-telemetry data
for other ungulate species. The key conclusions and recommendations from this research
project follow below.
Recommendations are highlighted using the symbol: 􀀩􂤶
6.1 PRONGHORN PERMEABILITY
• US 89 constitutes a near-total barrier to the passage of pronghorn, with only one
of 37 animals having crossed the highway in two years of GPS telemetry tracking.
• The pronghorn highway crossing rate averaged only 0.001 crossings/day among
30 pronghorn that approached US 89 to within 0.15 mile. The pronghorn passage
rate was negligible; only 0.006 crossings/approach.
• Due to the barrier effect and consequent few crossings by pronghorn, no
collisions with vehicles were recorded during the study, nor were any pronghorn-vehicle
collision records found in ADOT roadkill databases dating back to 1990.
• The barrier effect associated with US 89, coupled with the continued viability and
size of the pronghorn population points to the need for and potential benefit of
passage structures to promote permeability and maintain population viability.
6.2 POTENTIAL PASSAGE STRUCTURE LOCATIONS AND SPACING
• Based on weighted approaches by GPS-collared pronghorn to within 0.3 mile of
US 89, combined with several other factors analyzed at the 0.6-mile segment
scale, potential locations for pronghorn passage structures were objectively rated.
Pronghorn approaches to US 89 did not occur in a random manner, but rather
exhibited peak approach zones.
• Bissonette and Adair (2008) recommended spacing of 2.0 miles between passage
structures to accommodate pronghorn permeability based on theoretical isometric
scaling of home ranges. However, based on the empirical findings from this
project, the research team arrived at a different recommendation:
􀀩􂥔 The research team recommends a spacing of 3.2 miles between passage
structures which reflects: 1) the mean daily movement distance of
pronghorn, 2) half the mean linear approach distance by pronghorn along
US 89 determined over the duration of tracking, and 3) the “flattening”
effect of the highway on pronghorn home ranges that abut US 89.
50
• Based on the weighted pronghorn approaches and other rating factors, as well as
the research team’s spacing recommendation, three sites are recommended for
potential passage structures between MP 440.0 and 458.0.
􀀩􂥔 The highest priority for a passage structure was the section of US 89
between MP 440.6 and 441.1 (segments 136−141), the Coconino NF -
Antelope Hills site. This site had the largest peak in pronghorn
approaches and numbers of different pronghorn approaching the highway,
with 72% of GPS-collared animals on the west side approaching the
highway here, and 46% of the animals on the east side approaching it.
This site falls within a portion of US 89 that has already been
reconstructed to four lanes, but is close enough to the future reconstruction
stretch for passage structure construction be permissible here.
􀀩􂥔 The Wupatki National Monument passage structure site (MP 444.2−444.6;
segments 172−177), located three miles to the north of the Coconino NF-Antelope
Hills site, is situated amongst four relatively high rated segments
on Wupatki National Monument. This site is attractive due to the ease of
addressing ROW fencing issues since no livestock grazing occurs here,
and the planned highway median is considerably narrower (54 ft) than
other sites’ medians.
􀀩􂥁 A third passage structure is recommended at the Babbitt Ranch site (MP
447.2−447.7, segments 202−207), three miles north of the Wupatki
National Monument site. This site spans both private lands belonging to
Babbitt Ranches and State Trust land. The rating for this segment reflects
Babbitt Ranches’ proactive role in pronghorn management, including
further modification of ROW fencing.
􀀩􂥗 Where passage structures are considered, long-term land tenure must be
secure to ensure that the structures yield their intended benefit in
promoting pronghorn passage relative to the cost. Structures constructed
on USFS and NPS lands are secure, while strategies such as conservation
easements or long-term cooperative agreements may be needed on private
or State Trust lands to ensure similar long-term benefit.
6.3 IMPACT OF TRAFFIC AND NOISE
• Pronghorn distribution among all distances and across all traffic volumes up to
500 vehicles/hr remained constant. Only at 500-600 vehicles/hr (equivalent to
12,000−14,400 AADT) did a change in distribution occur. At volumes between
500 and 600 vehicles/hr, mean pronghorn distribution probabilities between
1,320−2,970 ft declined 17% compared to lower traffic volume classes, and
increased 73% in the 2,970−3,330-ft class. In the absence of attractive habitats
adjacent to US 89, pronghorn appear to lack an incentive to approach any closer.
51
• Traffic, at even low volume, consistently has a negative impact on pronghorn.
Daytime traffic volumes along US 89 typically exceed 10,000 vehicles/day, the
level at which highways become strong barriers to wildlife passage and traffic
repels animals from the roadway, as hypothesized by Mueller and Berthoud
(1997). Pronghorn are primarily active during daytime hours when traffic volume
peaks on US 89.
􀀩􂥁 A comprehensive set of measures to reduce traffic-associated noise impact
should be considered to create “quiet zones” along the highway at passage
structure locations to facilitate pronghorn highway approaches and
crossings. These measures could include recessing the roadway below
grade, integrating noise barriers such as berms, vegetation, and sound
walls, and applying pavement treatments like rubberized asphalt. Without
a comprehensive effort to reduce noise and visual impact, the success of
passage structures could be compromised.
6.4 ROLE OF FENCING
• Because pronghorn have exhibited limited ability to adapt to fences like other
ungulate species, ROW fences contribute significantly to the highway barrier
effect. And while fencing has been instrumental in funneling other animals to
passage structures to promote permeability, such an approach may not be
necessary for pronghorn because they seldom cross the highway. Rather, fencing
in conjunction with passage structures may be more useful in providing a visual
clue as to a path of least resistance across the highway barrier, provided no fencing
is used at the mouth of the passages.
􀀩􂥉 Ideally, passage structures should be located in areas with no ROW or
livestock pasture fencing near the highway as fencing presents an
impediment to free passage by pronghorn to and across the highway.
􀀩􂥗 Where ROW and livestock pasture fencing is needed (e.g., outside
Wupatki National Monument) to prevent livestock access to US 89,
creative approaches should be used to minimize the barrier effect of
fencing on pronghorn. Near passage structures, fences can be pulled
back from the highway ¼−½ mile to separate fencing and highway
barriers. A better approach still would be the long-term resting (or
temporary removal) of livestock pastures adjacent to passage structures
with the removal of fencing at the mouths of passage structures.
6.5 PASSAGE STRUCTURE DESIGN CRITERIA
• Site specific characteristics associated with different passage structure locations
will dictate what type of structure (e.g., underpass, viaduct, overpass) would be
appropriate from engineering and cost standpoints. However, structural design
characteristics will have a significant bearing on pronghorns’ eventual use and
acceptance of the passage structures.
52
􀀩􂥔 The most important structural consideration is the requirement that any
type of passage structure to promote pronghorn passage be as open and
wide as possible, with attention paid to avoiding obstructed line-of-sight
views through or across the structures or any restrictions to mobility.
• To date, no passage structure designed specifically to accommodate pronghorn
passage has been constructed in North America. As such, limited guidelines or
insights exist as to what type(s) of structure is best suited to promoting pronghorn
permeability. The research team believes that overpasses and/or large elevated
viaducts have the best potential for promoting permeability. Relative to the three
recommended locations for passage structures, the team recommends the
following structure types.
􀀩􂥔 The terrain at the Coconino NF-Antelope Hills site is suited to the
construction of an overpass for pronghorn passage. Since this stretch of
US 89 has been reconstructed, a retrofit application here is appropriate.
The application of CON/SPAN® pre-cast concrete arches may hold
potential for this site, as they can be dropped into place with minimal
traffic disruption, cost less than traditional bridges, and side-by-side
installation with arches of different heights could maintain existing
sloping ridgeline integrity. The headwalls could be flared to maximize
openness, with a 100-ft width to create a wide, open overpass.
􀀩􂥄 Due to the limited experience and insights on what structures will best
promote pronghorn permeability and whether they will successfully be
used along US 89 due to traffic impact and other factors, the research team
recommends that implementation of the Coconino NF-Antelope Hills
structure (MP 440.9) be considered prior to US 89 reconstruction under an
experiment