Wildlife issues are on the rise for state DOTs. Wildlife related concerns include habitat fragmentation and connectivity for wildlife, loss of habitat, increasing numbers of threatened and endangered species, and secondary and cumulative impacts.

The federal Endangered Species Act prohibits harm to or take of any listed species or adverse modification of designated critical habitat. Some of the existing regulations can be reviewed at the Overview of the U.S. Fish and Wildlife Service and recent developments at AASHTO's Center for Environmental Excellence website. Other sites which provide regulatory information include the USFWS' Digest of Federal Resource Laws of Interest to the U.S. Fish and Wildlife Service , Executive Orders, FHWA's Summary of Environmental Legislation Affecting Transportation, December 1998 and the Center for Wildlife Law' Federal Wildlife and Related Laws Handbook, Statute Summaries . Maintenance and construction staff are responsible for ensuring that no threatened or endangered species within areas they are working are injured or destroyed or their habitat impacted without proper permits. DOTs are implementing stewardship practices specific to certain threatened and endangered species as well as practices designed to benefit wider groups of species and ecosystems more proactively. This sections specifically focuses on what DOTs are doing to improve habitat connectivity and the ability of wildlife to safely cross roads.

State transportation agencies currently employ a mix of underpasses, bridge extensions, culvert installations, and culvert modifications, and associated fencing and ecowalls to facilitate wildlife movement. Effective wildlife fencing and crossing structures can significantly reduce many harmful impacts of roads on wildlife populations, though such measures can contribute to habitat fragmentation. More and more DOTs are exploring wildlife passages and culvert retrofits as means to enhance wildlife passage. In a few cases overpasses are being built. From a wildlife conservation perspective, the impacts addressed by these stewardship measures include habitat loss, habitat fragmentation, degradation of habitat quality, road avoidance zones, increased human activities, direct mortality, reduced biodiversity, genetic isolation, chemical contamination, changed hydrology for fisheries, reduced access to vital habitat, disruption of processes important to animal life cycles, and disruption of the food chain. Rigorous evaluation of the success of these measures has been implemented more slowly as funding for such work after the conclusion of construction is harder to find and frequently involves partnerships with others. In a 2002 survey by the author, 17 of 50 state DOTs said they were beginning to systematically incorporate wildlife crossings into roadway designs, but only a couple state DOTs had wildlife crossing policies, and only two states reported they had prioritized barriers for correction based on benefit evaluation. [N]

NCHRP 25-27, starting in 2004 and concluding in 2007, will investigate what guidelines are currently used by DOTs to determine wildlife crossing dimensions or design, the decision process, and any tools or aids that are used in that process. As of early 2002, only five state DOTs reported providing some direction to designers in this regard. In dissecting existing decision processes, NCHRP 25-27 will seek to understand the various factors used to decide what type of crossing will be employed as well as the extent to which long-term maintenance costs (annuities) of a highway structure guide selection (e.g. steel arch culvert vs. precast concrete girders with concrete deck). Ultimately the project will produce design guidance and a decision support tool for DOTs, as well as measures of cost and effectiveness. [N]

The locations of wildlife crossings and/or problem areas are critical for effective mitigation of the barrier effect caused by highways; however, few methodological approaches to identify and prioritize these key areas have been explored. Researchers from Madrid University found that 70 percent of collisions occurred on just 7.7 percent of the roads in the area they studied. [N] Collisions tended to occur where animals found it easier to cross roads in the absence of human habitation; fences or large, steep embankments deter animals from crossing roads and funnel animals to easy crossing points. Habitat availability on either side of the road is another factor in where animals choose to cross.

Only three state DOTs in the U.S. reported using modelling tools to identify habitat linkages as of 2002, and only one state did so for a large number of species on a statewide basis. [N] Six states had embarked on or participated in statewide efforts to determine connectivity needs as of early 2002. [N] Several more have begun to do so over the past two years.

Planning for wildlife crossings can be very involved; however, much can be accomplished using rapid assessment techniques, available information, and expert panels. The Forest Service's national expert on wildlife crossings advocates a simple rapid assessment approach that has proven effective in a variety of circumstances, is fast, and affordable: [N]

• Select highways to be examined.
• Select species for analysis
• Use available, spatially specific information, especially that available digitally in geographic information systems (GIS).
• Use a team of local biologists. Teams can often examine 100 miles/day.

DOTs, resource agencies, universities and non-governmental organizations such as The Nature Conservancy, American Wildlands, and the Rocky Mountain Elk Foundation have attempted to fill this gap by conducting workshops, often with DOT sponsorship. Biologists, researchers, and regulatory specialists come together in a workshop setting to make decisions on conservation and connectivity needs based on analysis of best available environmental data. With recent requirements of all states to identify priority conservation areas, new opportunities for interagency identification and prioritization of wildlife crossing needs are emerging.

NCHRP 25-27 will survey state transportation agencies to determine what information is used to determine location and number of wildlife crossings on planned sections of highway-improvement project, whether models are developed around political, project, or ecological boundaries, and whether connectivity needs are assessed at the project level, political/statewide level, or the level of ecoregions surpassing political boundaries.

Performance evaluation of crossing structure depends upon adequate monitoring. Future design benefits from such information as well. Most monitoring efforts to evaluate wildlife crossings have been short-term and focused on discerning whether target species are using the crossings. Monitoring programs have largely been aimed at single-species or have been confined to certain taxonomic groups; consequently, such programs may fail to recognize the requirements of other non-target species and ecological processes. Further, studies have generally failed to address the need for wildlife adaptation to environmental change. How well crossings ultimately perform depends on how well they accommodate changes in wildlife species distributions, abundance and behavioral profiles. Wildlife crossings are expensive measures, but a large void exists in devising cost-effective designs based on ecological and engineering criteria. Also, the current information base on wildlife crossing performance is geographically limited.

While nine of 50 state transportation had modified culverts for enhancement of connectivity across roads as of 2002, only four of these monitored such crossings with feedback to DOT designers and/or the state or federal wildlife agency. [N] Eight of 50 state DOTs reported monitoring specially constructed wildlife underpasses, with feedback to design and wildlife agencies. [N] Though DOTs more commonly employed bridge extensions to facilitate wildlife passage, usage by wildlife was much less frequently monitored than with other crossing technologies. [N]

According to research by the Western Transportation Institute, evaluation of a wildlife crossing structure installation may involve consideration of the following issues: [N]

• Motorist safety and animal-vehicle collisions
• Ecological impacts of mortalities and the "barrier effect" due to roads and traffic on individual animals, on a specific species, on populations of animals, on ecological communities and biodiversity, or on ecosystem processes and functional landscape integrity.

NCHRP 25-27 will add to this knowledge base by collecting details about types and methods of wildlife crossing monitoring in use, how often monitoring occurs, and the length of time for which it has occurred. The project will seek to detail the ecological criteria currently used to judge whether wildlife crossings are functional or effective, whether targets are established in advance, criteria are based on single target species, observed frequency of use by target or multiple species, population- or ecosystem-level data collection and analysis, or accident reduction.

This section reviews environmental stewardship practices and features that are being incorporated in many transportation projects across the country, including the restoration or preservation of habitat as mitigation and the addition of wildlife underpasses, overpasses, bridge extensions, enlarged culverts, and fencing by many states dealing with the realities of wildlife conservation and motorist safety. These practices were initially presented in NCHRP Report 305 on Interaction Between Roadways and Wildlife Ecology and are summarized herein. Habitat-related mitigation and conservation measures are some of the most effective measures and are used to address the broader ecological concerns associated with reductions in habitat and wildlife connectivity. Model stewardship practices in this area will also be briefly reviewed.

Existing crossing techniques can also be view at the Wildlife Crossing Toolkit developed by the U.S. Forest Service. Their website offers a searchable database of case histories from a wide variety of locations, time periods and wildlife species where people have attempted to solve issues resulting from wildlife/highway interactions. [N]

Research by the Western Transportation Institute found that the physical dimensions of the underpasses had little effect on passage because animals in the Banff National Park in Canada may have adapted to the 12-year old underpasses; however, structural attributes were more important on newer structures. Furthermore, the level of human activity in the vicinity was an important factor suggesting that mitigation strategies need to be proactive at the site and landscape level to ensure that crossing structures remain functional over time, including human use management. [N] The Western Transportation Institute is developing a vehicle-animal collision "toolbox" of countermeasures, which will provide detailed information to support application choices and decisions, and performance measurement. [N]

The remainder of this section consists of summary excerpts from NCHRP Synthesis Report 305 on Interaction Between Roadways and Wildlife Ecology and practices adapted from that discussion with regard to wildlife crossings, except where otherwise noted.[N]

• Design fencing applications for target species. Typical fencing applications are rectangular mesh or chain link fence from 2.6 to 3.0 m (8.5-10 ft) high. Florida and some European countries use strands of barbed wire along the top of the fence to discourage animals from climbing over the fence. Also used is finer mesh wire of from 2 x 2 cm (0.78 in.) to 4 x 4 cm (1.57 in.) buried 20 to 40 cm (7.87-15.75 in.) with a height extending from the ground of from 0.5 to 1 m (1.64-3.28 ft).
• Keep local wildlife interactions in mind when designing fences. For example, in California fencing application in areas with kit fox and coyotes provide a gap under the fence just large enough for the kit fox to negotiate at full run so that the latter can escape predators such as the coyote.
• For reptiles and amphibians, bend the upper edge of the finer mesh at a 90-degree angle to provide a lip to prevent animals from climbing over the fence. In Waterton Park, Canada, a temporary silt barrier type fence was used to direct frogs into polyvinyl chloride (PVC) drop traps so that volunteers could move them across the highway to a pond during the few-week-long migration period. Europeans have used a PVC barrier with an angled lip to keep reptiles and amphibians off the highways as well as a fabricated galvanized steel rail with a barrier lip along the upper edge. Iowa DOT has placed finer mesh fence at the bottom of regular fence to prevent smaller wildlife such as turtles, snakes and other small animals from getting on the Eddyville Bypass and Highway 63 at the Bremer-Chicksaw county line. This fencing approach has been commonly used in Europe to keep smaller animals off highways.
• Bury fine-meshed fencing at the bottom in use with pipe culverts for small animal connectivity or in association with other ROW fences. This practice has been successfully used under highways in Europe in culvert pipes with diameters approximately 0.4 m to 2.0 m (1.31-6.56 ft). California used a unique fencing application for desert tortoise approximately 6 km (3.7 mi) east of Kramer Junction on Highway 58 in San Bernardino County. A finer [1.27 cm (0.5 in.)] mesh section of wire fence, approximately 50.8 cm (20 in.) in height, was installed along the bottom of a typical 1.22-m (4-ft) right-of-way fence. The finer mesh fence was buried approximately 15 cm (5.9 in.) to prevent animals from going under. This portion of the fence was held in place using three strands of wire. The fencing application was done on an approximately 35.42- km (22-mi) section of four-lane highway. The fencing angled into the road at a series of culverts and bridges that were constructed for wildlife connectivity.
• Fence installation decisions should account for the potential for wildlife to be trapped between the fences should they find a way to enter the rights-of-way under, over, or around the fence ends. Because fencing is not totally exclusionary, Bissonette and Hammer studied two highway sites in Utah to compare the use of one-way gates and earthen ramps. They found that earthen ramps were used from 8 to 11 times more than one-way gates. Irrespective of the species, fencing without provisions for movement across the road can cause disruption of connectivity resulting in isolation of populations. This can be especially problematic for species with low populations, where the possibility of extinction can result.

Table 8 : U.S. Forest Service Wildlife Crossing Toolkit Guide to Culverts

Category	Definition	Typical Material
Box Culvert	Culvert has four sides, including bottom. Sometimes square or rectangular corrugated metal pipe culverts without bottoms are called box culverts, but in this toolkit they are referred to as bottomless culverts. Box culverts may be arranged in a horizontal series of small culverts to form multiple chambers.	Precast concrete Cast-in-place concrete Wood
Culvert (Continuous)	Culvert is continuous in circumference. The lower portion may or may not be buried. Sometimes simply called pipe. European badger culverts are sometimes called ecopipes. Slotted drain culverts are continuous except for a break in the upper portion.	Corrugated metal pipe Metal plate Cast-in-place concrete Precast concrete Wood
Bottomless Culvert	Culvert is discontinuous in circumference with rounded or square top and natural surface bottom. Also called open-bottom culvert.	Corrugated metal pipe Metal plate Precast concrete Cast-in-place concrete Wood

From Forest Service Wildlife Crossings Toolkit on-line at http://www.wildlifecrossings.info/cst2.htm

Modified Drainage Culverts
Drainage culverts can be modified to accommodate wildlife, a practice increasingly implemented by state DOTs and for which the Netherlands is renowned. Small mammals and amphibians are able to move through the culverts on shelves and floating docks or through wildlife tunnels built parallel to the wet culvert. NCHRP Synthesis Report 305 identifies the following stewardship practices with regard to modified drainage culverts: [N]

• Addition of a 0.46 m wide × 0.30 m high (18 in. × 12 in.) elevated concrete walkway can allow animals to move through a culvert even when water is present. Such modified culverts were successfully tried on US-98 in Texas for bobcats.
• A central "sacrificial" culvert with other culverts placed on both sides but at a higher elevation for drainage and connectivity can diminish blockage by beavers, which attempt to block the lower culvert, and the adjacent culverts at higher elevation remain open. Beaver-exclusion devices, including fencing, perforated pipe, or a combination of the two have been successfully used throughout the United States to reduce impounding of water behind road fills and for wetland mitigation and habitat preservation.

Colorado DOT has addressed a number of safety issues with beaver damage to roadways by using a modified culvert such as that described above. Safety issues have included potential to damage to the roadbed from water saturation, loss of flood capacity that otherwise would have been provided in the area flooded by the beaver dams, potential for heaving of the road way from frost or ice, and concern for motorist safety should they skid into the deeper water caused by the beaver dams. CDOT also wanted to cooperate with residents of nearby neighborhoods who desired the beavers to remain. To allow beavers to remain, CDOT perforated the beaver dams with pipes long enough to defeat the beavers natural inclination to plug holes in their dams. The pipes were inserted at a level that would allow the beavers to have a pond but not at such a level that the water continued to be a serious threat to the highway. Wire cages were placed over the ends of the pipes to prevent the beavers from clogging the pipes. [N]

Culverts with dual drainage and wildlife crossing purposes have proven successful in accommodating both terrestrial and aquatic organisms depending on water levels in the area of the culverts. When placed at the proper elevation, they can serve both types of organisms. Drainage culverts are typically used where highway causeways or fill sections transverse wetlands with fluctuating water levels such as wet prairies and marsh. They are also used on intermittent streams and floodplain areas that may inundate during wet periods. Aquatic species such amphibians and fish use them when they are wet and terrestrial species including reptiles and small mammals use them when they are dry.

Pictures of Florida DOT drainage/wildlife culverts, Massachusetts amphibian tunnels, Dutch structures, and wildlife using the culverts mentioned above, as well as other projects around the world can be seen at FHWA's wildlife crossings website; however, Evink notes that few states have researched the effectiveness of these structures. [N] Nevertheless, it is known that

• Culverts should be sized for use by multiple species, wherever possible.

• Use the natural stream bottom rather than a concrete or metal bottom.
• Provide shallow water or even dry edges along the stream edge in the culvert or bridge to allow the greatest number of species to move through.
• Allow extra height for larger mammals, such as deer, bear, and other species that ordinarily follow riparian corridors for movement and pass safely under roads. A wide variety of designs are possible depending on the site-specific construction environment – concrete box culverts, and round, oval, and elliptical pipe culverts.
• Provide cover as well as substrate on the inside of the culvert, similar to that of the exposed stream, to expand utility of the culvert. Proper sizing of the culvert depends on site-specific considerations and hydraulics, but including the natural streambed and as much adjacent upland as possible proves most successful.

• A range of culvert sizes, from 1.22 m × 1.22 m (4 ft × 4 ft) in Arkansas up to 2.44 m × 7.32 m (8 ft × 24 ft) in Florida and from 1.5 to 10 m (4.92–32.81 ft) in New South Wales, Australia, have been successfully used for various species of terrestrial mammals and reptiles.
• Use frequently placed culverts [150–300 m (492.13–984.25 ft)] of varying size in close proximity to shrub or tree cover. Clevenger and Waltho found that for a variety of culvert sizes for small and medium-sized mammals, passage was positively correlated with traffic density, road width, road clearance, and culvert length. They also found that all species with the exception of coyotes and shrews preferred small culverts with low openness ratios. There appeared to be some evidence of predation at crossings so the thought was that the smaller culverts provided comfort to the prey species. Weasels and shrews preferred culverts with nearby cover. The researchers felt that drainage culverts could be used to mitigate the harmful effects of high-speed roads. [N]

Wildlife underpasses are bridges and/or large culverts over dry land and sometimes land and water, constructed expressly to facilitate wildlife movement in important corridor areas. The length and height of these large culverts or bridges varies with the wildlife expected to use them. Twenty-three states report using underpasses for wildlife. The Western Transportation Institute is cataloging existing uses of and research on underpasses and culverts to provide further guidance to DOTs on stewardship practice in this area. [N]

• Provide adequate area for both water movement with associated organisms and dry habitat for terrestrial species movement.
• Consider the characteristics of the area when trying to determine the appropriate length of the bridge. In cases where there is an important corridor for movement of rare or protected wildlife species, bridging the entire floodplain may be necessary. At the other end of the spectrum, where the floodplain is being used by habitat-limited species, a combination of smaller structures and fences may be possible.
• When choosing a combination of bridge and fill, consider what reptile and amphibian species will likely move up the fill slope onto the road. Standard fencing will not stop this movement so that very expensive barrier walls and associated guard rails may be necessary to prevent significant kills of these species during periods of the year when they are moving around in large groups.
• Consider the cost of mitigation for wetland takings by opting for a fill section. By the time the costs of shorter bridges or culverts, fill acquisition, barrier walls for reptiles and amphibians, guardrails, and fencing are factored in, along with the cost of wetland mitigation, the cost of a more substantial bridge, preferred for habitat connectivity, may already have been approached.

Viaducts
Viaducts are a potential solution for the entire spectrum of species moving through an area, as these long bridges can leave wetlands, rivers, and variable topography and geology below largely intact. Typically, this approach is most cost-effective where there is topographic relief, such as in mountainous areas, is sufficient to make bridging necessary for a significant span of a waterway, canyon, or valley; use of spanned lands by wildlife is typically a secondary benefit. [N] Design of viaducts for wildlife connectivity and habitat enhancement is increasing, especially in Europe.

Wildlife Overpasses
Although wildlife overpasses are largely a European phenomenon, Florida, Hawaii, New Jersey, and Utah reported overpasses being used by wildlife and the Montana Department of Transportation and the Connecticut DOT are installing wildlife overpasses on Highway 93 and Route 6, respectively. The New Jersey overpasses, among the first in the United States, were completed in 1985 at a cost of $12 million. The overpasses were designed to provide connectivity across I-78 (a six-lane highway) at an approximately 2-mile stretch that crossed the Watchung Reservation in Union County. Specifically, one was constructed solely for wildlife use (especially deer). Another one was constructed for shared wildlife and vehicular use; however, it has since been closed to traffic. A third overpass was constructed as a shared use between wildlife, vehicular traffic, and a bridle path for horses. The Utah overpass was constructed principally for deer. The Florida overpass on I-75 in Marion County just north of County Road 484 is a multi-use overpass designed to accommodate a recreational and equestrian trail, as well as for wildlife use. Two overpasses were built over the Trans-Canada Highway in Banff National Park and are being used by a variety of wildlife. [N]

• Allow the widest width possible. Wildlife overpasses can vary in width from 3.4 m (11.15 ft) to 870 m (2,853.6 ft). Wider passages are more effective at allowing animals to cross and animal behavior on wider structures is more normal than on narrower ones.
• Establish or preserve suitable habitat at and leading to the overpasses. Where this has been accomplished, it has been found that the overpasses were effective for a wide variety of animals including invertebrates.

• Use a series of solar-powered, battery-operated, motion sensors to determine animal presence and trigger low-voltage, LED-illuminated warning signs that reduce the posted speed limit to 40 km/h (25mi/h) and alert motorists to the presence of approaching wildlife. This method has been successfully used in Switzerland, though the location on the road was also adjacent to a large wildlife overpass over a major nearby freeway.
• Install vertical pipes perpendicular to bridge railings to keep bird flight patterns above the elevation of traffic. Florida reported installing PVC pipe approximately 3 m (9.84 ft) in height perpendicular to the railing on the San Sebastian Bridge. The poles were spaced approximately 3.7 m (12 ft) and kept birds hovering over the bridge from dropping down into traffic crossing the bridge, reducing bird kills.

Most state DOTs try to find a land management agency to provide maintenance for mitigation/conservation habitats. A few have sought to perform mitigation (modification or restoration) on public lands that are already being managed by a resource agency. While some federal resource agencies encourage this approach, others disallow it. Universities, conservation groups, resource agencies, and even private groups where consistent with the objectives of the mitigation have become involved in maintenance of habitat by taking possession or easement of land from the state DOTs. Few transportation departments are maintaining habitat except for wetland mitigation sites, though creative conservation partnerships have been developed in a number of states through in lieu fee arrangements. Texas, Tennessee, Kentucky, North Carolina, and Florida are among the DOTs allowed to provide funding instead of mitigation. Habitat management is frequently the responsibility of the agency or partnering organization holding title to the land or conservation easement.

In most cases, when a state uses a habitat strategy as a conservation or mitigation measure, the maintenance plans for these habitats contain a monitoring requirement. These requirements vary in length and design, but states are often required to maintain sites to varying performance levels (such as a percent survival of desired species and exotic/invasives free) for some period of time (commonly 3 to 5 years). Specific management plans including funding can also be a requirement. Such a monitoring program is designed to:

• Specify recommended mitigation and ensure that it is included in the final design process.
• Monitor the implementation of the mitigation through design, construction, and operation.
• Resolve issues that are contingent on the outcome of design as it progresses to more detailed stages.
• Report on progress toward implementation of mitigation measures to responsible parties.

Biologists from resource and conservation agencies associated with CDOT's Shortgrass Prairie Initiative crafted recommended Baseline/Annual Reporting Requirements that would provide an effective yet practicable framework for adaptive management and annual reporting: [N]

• Type of plant communities/habitats present:
• Size/extent of each present plant community or habitat group.
• Description of plant communities present and estimated percent cover.
• General condition of each plant community/habitat type (estimated percent weed cover, estimated weed sp. relative abundance, estimated percent bare ground, etc.).
• Juxtaposition of communities/habitats.
• Brief description of land use on-site and in surrounding areas documented in initial baseline. Report changes over the past year.
• Recommendations for management of parcel (i.e., grazing, controlled burns, plantings, etc.) to achieve conservation goals.
• Success of recommendations from previous year and suggested modifications.
• Observations on wildlife diversity, activity, and general trends. i.e. field notes. Surveys and quantitative data are not required.
• Photo points at established permanent locations according to protocols to be developed in the management plan. The on-site managers and regulatory oversight agreed that new aerial photos would be acquired as they become available.