Evaluation of the Use and Effectiveness of Wildlife Crossings (2008)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

118 Top Five Priorities by Nation Tables 37 and 38 show the top five priorities of the United States and Canada for restoring wildlife movement across roads. Top Five Priorities of Engineers/ Analysts/GIS Specialists Table 39 shows the top five practice and research priorities of engineers, analysts, and GIS specialists. Top Five Priorities of Planners Table 40 shows the top five research and practice priorities of planners. Top Five Priorities of Natural Resource Professionals—Overall Table 41 shows the top five research and practice priorities of natural resource professionals. Plan of Action for Priorities Practice Gaps and Priorities Ecological Agencies responsible for creating wildlife mitigation measures along transportation corridors would profit by standardizing and institutionalizing practices that aid in the development of mitigation techniques. There is a need to standardized methods for collecting and recording data, the development and communication of state and provincial wildlife habitat conservation needs, and the development of wildlife crossing guidelines on state- and A P P E N D I X A Priority Tables and Plan of Action province-wide and regional bases. When transportation pro- fessionals and scientists seek readily available databases or systematic methodologies for gathering data so that they can incorporate ecological data into transportation programs or mitigation models and measures, they are apt to find that long-term databases, such as those containing wildlife–vehicle collision data are usually not associated or linked with spatially accurate locations, road geometrics, or environmental data. Additionally, databases may be in a variety of formats. For example, wildlife–vehicle collisions data are contained in highway patrol reports, and in forms filled out by highway maintenance crews. Important data such as the location and maintenance schedules for culverts and bridges are not always electronically available. There are also inconsistencies among states and provinces concerning the dissemination of critical wildlife habitat needs, and the identification of priority areas for conservation. If these data were readily available in electronic databases similar to each state’s Natural Heritage program but in greater detail, DOTs and MoTs would be better able to incorporate wildlife and ecosystem priorities into the planning stages of transportation programs and indi- vidual projects. Additionally the research team found that guidelines for planning and installing wildlife crossings are nonexistent for most states and provinces. In summary, the research team have found a lack of (1) long-term and accurate databases on wildlife–vehicle collisions or roadkill carcass locations that are electronically based and standardized, (2) a nation-wide standardized method for state and provincial wildlife agencies to incorporate wildlife locations and their habitats and needs in a cohesive document readily available for other agencies to work with, and (3) widely available guidelines for developing and maintaining wildlife crossings and other mitigation measures. Without standardized and institutionalized practices informed by accurate, complete, and documented databases, transportation professionals find it difficult to collect and analyze data on wildlife–vehicle col- lisions and roadkill carcass locations, include ecological and

119 United States Top Priorities Rank Canada Top Priorities Incorporate wildlife mitigation needs early in the DOT/MoT programming, planning, and design process 1 Same Combine animal-friendly mitigation methods such as wildlife crossings, fences, escape ramps, and gates, rather than using one method 2 Same Use conservation plans and connectivity analyses to inform the transportation programming/planning/design process on where mitigation is needed and how it may be carried out 3 Same Establish effective communication and collaboration among stakeholders 4 Use standardized and vetted protocols for collecting and recording roadkill carcass and wildlife–vehicle collision data Incorporate into plans and schedules wildlife crossing options that can be accomplished by maintenance crews simply by retrofitting existing facilities 5 Establish effective communication and collaboration among stakeholders Table 37. Top five priorities for restoring wildlife movement across roads in the United States and Canada. United States Top Priorities Rank Canada Top Priorities Understand better the dynamics of animal use of mitigation structures (such as what works and what does not) and disseminate this information 1 Same Develop and summarize alternative, cost- effective wildlife crossings designs and the principles they are based on 2 Standardize spatially accurate roadkill carcass and wildlife–vehicle collision data collection Develop wildlife crossing designs and guidelines for the full suite of animals in an area to help facilitate permeability for many species 3 Develop and summarize alternative, cost- effective wildlife crossings designs and the principles they are based on Develop state-based habitat connectivity analyses for every state 4 Develop guidelines to decide when wildlife mitigation is necessary (both mandated and voluntary) Develop a standardized monitoring protocol to assess crossing effectiveness 5 Develop prototype animal–vehicle collision safety models to predict where wildlife–vehicle collision hotspot areas are and may be on future roads Table 38. Top five research priorities for restoring wildlife movement across roads in United States and Canada.

120 safety data into the planning process, create mitigation meas- ures for wildlife, or find ways to integrate existing mainte- nance and upgrade schedules with mitigation opportunities. Priority. Create standardized protocols for collecting roadkill carcass locations and wildlife–vehicle collision data. This information is crucial in helping to determine where wildlife mitigation measures to reduce wildlife–vehicle collision are needed. Departments and Ministries of Trans- portation would benefit from the collection of data by standardized, accurate methods that could be incorporated with future road improvements, road building, and reduc- tions in wildlife-related crashes. (See priorities above under safety models). The research team points to two successful efforts: Maine has a wildlife–vehicle collision reporting program that is geo-referenced and mapped for the public,156 and British Columbia has maintained a long-term database of wildlife–vehicle collisions that is analyzed in order to create appropriate measures to reduce these crashes.211 Priority. Create continent-wide guidelines and standards for determining when during the transportation planning process agencies should assess programs and projects for wildlife needs. The U.S. Transportation Equity Act for the 21st Century (TEA-21) requires that planners develop long-range plans and short-range programs that consider projects and strategies that, among other things, will protect and enhance the environment. However, the act provides no guidance on how planners should meet these requirements.235 Typically, if ecosystem and wildlife are considered, it is late in the devel- opment of a transportation project. This can often lead to de- lays in the permitting process, incurring the expenditure of additional funds. This is not in the best interest of the ecolog- ical resource, the transportation agencies, or the public. The research team strongly suggests these analyses be incorporated early in the development of long-range transportation plans. Planners in Oregon, South Dakota, Colorado, and North Carolina for example, extensively consider ecosystem conser- vation during planning processes.235 Vermont has a policy of addressing wildlife and fish needs in future transportation projects prior to regulatory intervention (C. Slesar and J. Austin, personal communication) and Montana state and federal agency personnel have worked together to create the largest, most comprehensive sets of wildlife mitigation meas- ures over one highway in the United States.205 This priority is linked with the priorities to implement statewide connectivity Rank Practice Priorities 1 Incorporate wildlife mitigation needs early in the DOT/MoT programming, planning, and design process 2 Establish effective communication and collaboration among stakeholders 3 Combine animal-friendly mitigation methods such as wildlife crossings, fences, escape ramps, and gates, rather than using one method 4 Use conservation plans and connectivity analyses to inform the transportation programming/planning/design process on where mitigation is needed and how it may be carried out 5 Incorporate into plans and schedules wildlife crossing options that can be accomplished by maintenance crews simply by retrofitting existing facilities Research Priorities 1 Understand better the dynamics of animal use of mitigation structures (such as what works and what does not) and disseminate this information 2 Develop and summarize alternative, cost-effective wildlife crossings designs and the principles they are based on 3 Develop guidelines to decide when wildlife mitigation is necessary (both mandated and voluntary) 4 Develop a standardized monitoring protocol to assess crossing effectiveness 5 Develop wildlife crossing designs and guidelines for the full suite of animals in an area to help facilitate permeability for many species Table 39. Top five research and practice priorities of engineers/analysts/ GIS specialists.

121 analyses, enacting policy to mandate these actions and the pri- ority below, the development of statewide wildlife habitat conservation plans. Priority. Incorporate state- and province-wide maps and conservation plans for critical wildlife habitat needs into transportation planning. The U.S. Fish and Wildlife Service asked state wildlife agencies to complete their wildlife habitat conservation plans by October 2005. These plans will include GIS-generated maps showing ranges and critical habitats of species of concern and will greatly assist DOTs and MoTs in planning mitigation to maintain or restore ecosystem integrity and viable wildlife populations. The successes of current programs may be used to help guide further actions. For instance, Florida’s Fish and Wildlife Commission has communicated wildlife needs through its Integrated Wildlife Habitat Ranking System,76 the “Closing the Gaps in Florida’s Wildlife Habitat Conservation System” documentation, its statewide “Strategic Habitat Conservation Areas” program, and the “Biodiversity Hot Spots,” which are all scientific efforts that are translated into GIS data layers and are incor- porated into Florida DOT’s Environmental Screening Guide (V. Sharpe personal communication).92 Priority. Create and update guidelines for considering, placing, designing, and constructing wildlife crossings. This is a priority for the practice as well as the science of road ecology. This priority is linked to the development of monitoring pro- grams to assess the effectiveness of mitigation measures.196 More detailed statements related to this priority can be viewed above under guidelines in research priorities. Examples of current guideline efforts include the set of guidelines created for the installation of amphibian and reptile tunnels in New England,130 standards created for river and stream crossings Rank Practice Priorities 1 Combine animal-friendly mitigation methods such as wildlife crossings, fences, escape ramps, and gates, rather than using one method 2 Incorporate wildlife mitigation needs early in the DOT/MoT programming, planning, and design process 3 Use conservation plans and connectivity analyses to inform the transportation programming/planning/design process on where mitigation is needed and how it may be carried out 4 Incorporate into plans and schedules wildlif e crossing options that can be accomplished by maintenance crews simply by retrofitting existing facilities 5 Establish effective communication and collaboration among stakeholders Research Priorities 1 Understand better the dynamics of animal use of mitigation structures (such as what works and what does not) and disseminate this information 2 Develop and summarize alternative, cost-e ffective wildlife crossings designs and the principles they are based on 3 Develop wildlife crossing designs and guidelines for the full suite of animals in an area to help facilitate permeability for many species 4 Develop standardized inventories of wildlif e crossings by state for better management and maintenance of these crossings, and to better assess the need for future crossings 5 Three priorities tied for fifth rank: Develop a standardized monitoring protocol to assess crossing effectiveness Develop guidelines to decide when wildlif e mitigation is necessary (both mandated and voluntary) Increase our understanding of the effects of road density on wildlife populations Table 40. Top five research and practice priorities of planners.

122 Rank Practice Priorities 1 Incorporate wildlife mitigation needs early in the DOT/MoT programming, planning, and design process 2 Combine animal-friendly mitigation methods such as wildlife crossings, fences, escape ramps, and gates, rather than using one method 3 Use conservation plans and connectivity analyses to inform the transportation programming/planning/design process on where mitigation is needed and how it may be carried out 4 Establish effective communication and collaboration among stakeholders 5 Incorporate into plans and schedules wildlife crossing options that can be accomplished by maintenance crews simply by retrofitting existing facilities Research Priorities 1 Understand better the dynamics of animal use of mitigation structures (such as what works and what does not) and disseminate this information 2 Develop and summarize alternative, cost-effective wildlife crossings designs and the principles they are based on 3 Develop wildlife crossing designs and guidelines for the full suite of animals in an area to help facilitate permeability for many species 4 Develop state-based habitat connectivity analyses for every state 5 Develop a standardized monitoring protocol to assess crossing effectiveness Table 41. Top five research and practice priorities of natural resource professionals. for fish,236 and Colorado’s guidelines for the placement of crossing opportunities for wildlife.14 The United States has learned and will continue to learn from one of the agency lead- ers in wildlife crossings, Parks Canada, who has taken the lead for the North American continent in instituting and evaluat- ing wildlife crossings. Priority. Funding and maintenance for an outlet that communicates the standardized guidelines is a priority. Guide- lines could be communicated and updated through the use of agency-based websites. As knowledge improves and evolves, so should the guidelines be updated to reflect this or they risk be- coming obsolete. There is a need for meaningful partnerships among federal agencies and associations to commit resources and personnel to maintain useful websites. Ideally, these sites should provide extensive searchable literature databases, annotated bibliographies, research reviews linked to projects, as well as the previously mentioned guidelines, and any deci- sion guide associated with NCHRP Project 25-27. Linking to existing websites (i.e., the Wildlife Crossings Guidekit, the Deer–Vehicle Crash Information Clearinghouse) would be most helpful. This is an ongoing priority that needs the con- tinued guidance and attention from a multi-agency committee that has credibility with transportation professionals. Priority. Maintenance activities on crossings need to be recorded in standardized documentation schedules. If struc- tures and accompanying mitigation features such as fences are not maintained, their effectiveness often decreases.71 Documentation of maintenance schedules, methods, and costs provides assurance that the structures are fulfilling their purpose and can help in establishing maintenance needs for future mitigation measures. Priority. Create alignment specifications for effective mit- igation efforts that link wildlife crossings with fences and right- of-way (ROW) escape ramps. Certain fence types are known to not be wildlife friendly, and wildlife need to be able to escape if trapped on the roadway. Evidence from studies conducted at Utah State University suggest that the rectangular mesh design used in most “deer-proof” fencing applications can result in the death of juvenile animals who become trapped in the fence.

123 The barbed wire arrangement used in lower fences is also prob- lematic for species such as pronghorn antelope. The research team urges standards for animal-friendly fences, for example, a different mesh size for exclusion fences. Larger animals often access the ROW even if it is fenced. Measures (such as earthen escape ramps) are needed to allow them to escape the road ROW in the presence of exclusion fencing. Fencing mitigation efforts need to incorporate escape ramps in order to be maxi- mally effective. Ungulate (deer, elk, and moose) are much more inclined to use escape ramps than “squeeze-through” steel gates to escape the ROW. Additionally, in areas with rugged topog- raphy, the typical perpendicular ramp-fence alignment may not be most effective. Ramps placed “in-line” with the fence may be a desired alternative. There is a need to explore other es- cape mechanisms that could be created for large and small an- imals, e.g., badger and small-mammal tunnels. The research team also encourages the practice of implementing alternative and innovative designs. These could be developed in an adap- tive management context of learning from doing; a context where practice is tied to research in an explicit fashion. This ap- proach is currently in use in developing wildlife mitigation measures in Arizona.73 For small animals that are not deterred by exclusion fences, the research team suggests the adoption of jersey barriers with wildlife scuppers (openings in the barrier that allow for passage of small animals and water movement) or low barriers that direct animals to small tunnel-like passages. Additionally, research has shown higher roadkill levels often occur at the end of the fenced mitigation, the so-called “end- of-fence” problem.71 The research team considers it a priority to address fence designs, the end-of-fence problem, and ramp- fence alignments in order to increase the effectiveness of these common mitigation structures. Priority. Culvert and bridge maintenance schedules need to be made available in electronic format so upgrade and replacement projects can be coordinated with mitigation measures. Existing transportation infrastructure could be retrofitted for wildlife and fish during routine maintenance and upgrading. Regional protocols could be developed to in- tegrate culvert, bridge, and fencing maintenance schedules with the needs of aquatic and terrestrial wildlife in the area. Protocols that retrofit culverts for fish passage are available in several states.16,157,164 As more is learned about efforts by states and provinces to create standardized collection methods and data storage, and to create guidelines for wildlife mitigation measures, the re- search team believes these efforts can be implemented across the continent. Policy and Planning National, state, and provincial authorities at the highest levels need to be fully engaged if policies and guidelines that mandate the use of standardized and effective methods to maintain and promote permeability and connectivity of the landscape for wildlife are to be enacted. If transportation and natural resource agencies continent- wide are to address the pressing issues of landscape fragmen- tation and effects of road transportation networks on species, it is essential to go beyond the individual transportation proj- ects and individual species approaches of the past.98 There is a need for national-level, firm, and legal guidelines that man- date the incorporation of wildlife and ecosystem considera- tions early in the long-range transportation planning stage. There is also a need to correct the basic inconsistencies among states and provinces in their practices and policies to- ward protecting wildlife and re-establishing connectivity across the landscape. To this end it is necessary to establish common goals and objectives that state/provincial and fed- eral governmental agencies can agree upon and accomplish in order to increase permeability of transportation corridors for wildlife. Priority. Legislation that enables and funds mandatory planning and mitigating for wildlife along transportation corridors is desirable. The research team believes this is attainable and point to two currently successful programs: Florida and The Netherlands. In both places, laws and poli- cies have been passed and programs funded to develop maps of ecological networks and to identify places where roads fragment or fracture these networks, and where specific areas and transportation projects for mitigation and compensation have been identified.11,94 In order for similar actions to be applied across the United States and Canada, there is a need for fully funded federal-level mandates or strong incentive programs that authorize and institutionalize methods to identify, plan, and mitigate for landscape connectivity along transportation corridors. Funding for this effort may be attained from TEA-21, FHWA research funds, and dedicated state funds.11,79 The research team suggests that leadership for these efforts to coordinate multi-agency standards that help maintain and promote permeability and connectivity of the landscape for wildlife come from a strong federal-state/provincial partner- ship. Likely partners include the U.S FHWA and Transport Canada as well as the U.S. Fish and Wildlife Service and Envi- ronment Canada, coupled with transportation and wildlife representatives from the states and provinces. Additional organizations might well include the American Association of State Highway and Transportation Officials (AASHTO), the U.S. Forest Service, the U.S. Environmental Protection Agency, and the Bureau of Land Management.11 Non-governmental organization (NGO) participation would provide a public input, and several have been active in this arena (Defenders of Wildlife, the BC Conservation Foundation, and The Nature

124 Conservancy). A successful partnering arrangement might partition duties with wildlife and natural resource agencies providing ecological guidelines as well as measures of success, while the FHWA and Transport Canada could standardize and coordinate compliance with state and provincial DOTs/MoTs, while at the same time linking incentives with funding. The research team strongly encourages funding incen- tives in these mandates for transportation agencies to (1) conduct connectivity studies, for example, as Washington State DOT has done 212; (2) fund GIS data development (which many DOTs/MoTs have done); (3) continue data- partnering with other agencies, which most states do but which can be improved 91; (4) incorporate wildlife habitat connectivity maps in long-range transportation program planning such as Washington and Florida have done; (5) mitigate for wildlife in most transportation projects (e.g., such as what is done in Vermont); and (6) remain consistently committed to this goal over time. Such a specifically defined approach would help to level the current inconsistencies among states and provinces and promote continent-wide permeability for wildlife across transportation corridors. Communication Improved communication among transportation pro- fessionals, on-the-ground transportation workers, scien- tists, activists, and the public is needed to help ensure that wildlife crossings measures and other actions to maintain ecosystem permeability across transportation networks are driven by the most effective and efficient methodologies. Road ecology is a rapidly developing field of study that re- lies on communication among researchers and practitioners from around the world. While great gains have been made in this field in the past 10 years, the research team has observed common gaps in communication within and among agencies and other professionals and the public that directly affect the ability to place and maintain wildlife crossings. It is impera- tive that study results and practices that are both successful and unsuccessful are communicated as quickly and effectively as possible in order for transportation planners and engineers to build transportation corridors that are less damaging to wildlife. Although not universal, the lack of communication and data sharing among agencies at federal and state levels that relate to long-range planning; the collection of accurate roadkill and wildlife–vehicle collision locations; and the effective placement, construction, monitoring, and mainte- nance of mitigation structures hampers progress. Addition- ally the lack of funding to coordinate such communication is problematic. The need for communication has been a common theme in the priorities listed in this document. The research team finds it is a key component to successful wildlife mitigation programs. The most successful and far reaching wildlife- transportation mitigation programs across the United States and the world have communications networks (such as the Infra Eco Network Europe [IENE]) that have been developed to coordinate information,91 include ecosystem-level needs in transportation planning (IENE’s Cost 341 effort),129 and garner support for providing measures for wildlife in trans- portation systems.11 From these and other examples of suc- cesses within states and provinces, the research team describes in the following paragraphs two priorities that relate to infor- mal and formal communication. Priority. Increased informal communication opportu- nities among transportation professionals, on-the-ground transportation workers, and ecologically trained profes- sionals are necessary. Some notable successes include the International Conference on Ecology and Transportation bi-annual events and the well-circulated proceedings from those conferences; the Center for Transportation and the Environment at North Carolina State University and its well- maintained website and list server; the Wildlife Crossings Guidekit website initiated by the U.S. Forest Service and housed at Utah State University; the Deer-Vehicle Crash Information Clearinghouse at the University of Wisconsin; and other events, publications, and websites dedicated to highlighting and exploring wildlife and transportation issues. The research team encourages transportation and wildlife professionals to communicate using the above-mentioned methods and other less formal means to learn about ecologi- cal impacts of roads, successes and failures of research and practices, and innovative ideas and for increased opportuni- ties to include wildlife and ecosystem needs into the planning and designing of roads. Increased communication opens opportunities to coordinate mitigation for ecosystem and wildlife needs in the development of long-range programs and project plans long before these plans became fully developed and budgeted. There is a strong need for direct communica- tion between biologists and on-the-ground transportation workers. These workers are the critical link to accurate data collection and are often the source of innovative design solu- tions. They are often very interested in wildlife and would like feedback on the effectiveness of mitigation measures that they design and install. They can also provide crucial information to biologists such as maintenance schedules for bridges, cul- verts, and upgrades to roads. If these schedules were coordi- nated with environmental managers and biologists, ideally these already planned projects could present opportunities to retrofit these structures for the movement of wildlife, fish, and ecological processes. With over 575,000 bridges in the United States and as many as 40,000 of these needing repair or replacement in the next two decades,253 there are literally tens

125 of thousands of opportunities to coordinate such efforts to improve landscape connectivity. Priority. Increased formal communications among states, provinces, and countries are necessary. The research team believes these communications are necessary to help this field move forward concerning the development of ef- fective mitigation structures. There are several avenues for increased communications including clearinghouses, con- ferences, proceedings, publications, and federally sponsored websites such as FHWA’s Exemplary Ecosystem Initiatives and Wildlife Protection—Keeping it Simple, the TRB’s Transportation Research Information Services database, AASHTO’s Center for Environmental Excellence, and Standing Committee on the Environment, British Columbia Conservation Foundation’s Wildlife Vehicle Accident Prevention Program, and Parks Canada’s Highway Mitiga- tion Research Program. The research team suggests a clearinghouse for projects across North America. This central location could be maintained by the FHWA and would house information on past, current, and future proj- ects with specifics that would be of interest to other agencies and locations. The research team suggests additional oppor- tunities to share information over the entire continent for example increasing the number of public meetings such as ICOET and regional ecology and transportation confer- ences, for instance the Northeast Wildlife and Transporta- tion Conference. The proceedings of these meetings are a major source of information on developments in this field. The proceedings and other information could be published in a way that professionals from a variety of non-ecological transportation interests such as planners, administrators, and engineers, would be notified electronically of their exis- tence. These proceedings and other publications would help promote the science base of road ecology if they gave easily accessed sets of definitions for all professionals to under- stand. Communication could also be improved if long-term funding was available to maintain websites dedicated to wildlife crossings and other related mitigation measures. Linking them to the FHWA website is a step in the right direction. Finally, professionals have a responsibility to educate and help the public become aware of issues concerning wildlife mortality, crossings, landscape frag- mentation versus permeability, and public safety. The research team encourages transportation agencies to com- municate with the public the needs for wildlife crossings projects, the development and completion of mitigation measures, and the results of monitoring projects. In these communications, the research team strongly encourages scientific messages pertaining to the issues of fragmentation and connectivity to help the public understand. The re- search team encourages progress in all these areas in order to quickly and efficiently bring about change in the practices associated with transportation and wildlife. Research Gaps and Priorities Safety Existing wildlife–vehicle collision prediction models re- quire further development to be effectively used for safety analyses tasks such as identifying wildlife collision-prone locations on both existing roads and new roads, evaluating the collision reduction effectiveness of mitigation meas- ures, and conducting cost-effectiveness analyses of poten- tial mitigation projects. There is a need for the development of more reliable wildlife–vehicle collision prediction models that would in- form transportation professionals about collision-prone areas, not only on existing roadways, but also on new road- ways in the planning or design stage. These models would assist in systematic screening of the road network, which is routinely done in jurisdictions, in order to identify specific lo- cations that merit further investigation as potential locations for crossing, fencing, and other mitigation measures such as those that address driver behavior (e.g., reduced speed limits). These same predictive models are also required to assess, retrospectively, the collision reduction effectiveness of coun- termeasures aimed at reducing wildlife–vehicle collisions. The types of models required for these purposes ideally would estimate the expected frequency of collisions. Most current site- specific models estimate the probability of a site being a “high- collision location,” which is subjectively defined, and therefore does not provide an estimate of the expected collision fre- quency.15,88,124 These “probability” models typically include variables that necessitate field data collection and thus they cannot be applied for network-wide screening because of data limitations in state databases.169 Additionally, most current wildlife–vehicle collisions prediction models are limited in their ability to accurately describe the general cause-and-effect relationships among variables that affect collisions and hence are limited in their ability to inform practitioners who would like to be proactive in predicting where wildlife–vehicle colli- sions are most likely to occur. The development of integrated models is hampered by (1) the lack of a national protocol for collecting wildlife–vehicle collision as well as roadkill carcass data; (2) the limited number of reliable long-term databases of wildlife–vehicle collisions and roadkill carcass data; (3) the lack of crash site data or other important model inputs such as highway variables (geometrics) and ecological variables (e.g., topography and existence of migration routes); and (4) the lack of knowledge of wildlife exposure (i.e., the change over time of the number or density of animals in enough prox- imity to a road to be potentially struck by a motor vehicle). It is

126 apparent (to the research team) that spatial accuracy is a defin- ing characteristic of these databases. The research team believes tremendous progress can be made in this research area if the following priorities can be accomplished. Priority. Develop a strategic plan that is a well defined, in- terdisciplinary, and multijurisdictional strategy to address the wildlife–vehicle collision problem and its complexities.142 There are dozens of attempts to model wildlife–vehicle collisions with different methods, in different regions over many different sit- uations, and yet the approaches tend to be piecemeal rather than building on one another. In order to bring the develop- ment of wildlife–vehicle collision predictive models to a level where they are applicable over large regions, the combined ef- forts of professionals in several disciplines is desirable. The re- search team believes that the current pooled-fund proposal 142 for the creation of a Deer-Vehicle Crash Information and Re- search Center is a step in the right direction for bringing the past and future work together in one central location and for the de- velopment of a cohesive strategy to address this issue. Priority. Standardize and improve the collection of roadkill carcass data and wildlife–vehicle collision data. Data on roadkills and wildlife–vehicle collisions are currently collected by a variety of methods. The research team knows of only one database that has the spatial accuracy needed to produce reliable ecological models that link environmental variables with road mortality of animals (the Parks Canada database). The research team suggests a roadkill collection protocol and a wildlife–vehicle collision location protocol be standardized across the nations or within regions in order to obtain spatially accurate reliable databases not only for modeling efforts, but also to assist in state Departments of Transportation (DOTs) and provincial Ministries of Trans- portation (MoTs) efforts to reduce collisions and roadkill. Data collection on collision and carcass sites could provide more accurate information if they were geo-referenced, i.e., identified with global positioning systems that accurately specify the collision–carcass location. These accurate loca- tions are critical if the research team is to assess the entire suite of other factors believed to affect collisions. Priority. Include spatially accurate information on off- roadway variables into highway safety models used to predict wildlife–vehicle collisions. If these data are not available, safety models could only be developed with only information pertaining to roads (road geometrics and traffic volumes). Such models, though still useful as predictive models, are lim- ited in their ability to advance understanding and capability to predict where and when wildlife–vehicle collisions will occur. Off-road information that would be considered in a model include variables known to affect wildlife movement across roads, such as presence of nearby fencing, culverts and bridges, presence and characteristics of wildlife underpasses, adjacent land cover, distance to cover from the edge of the road, topography, human use of the area, species present, and standard road geometrics. The research team believes that assembling information on variables such as these would provide much improved databases that could in turn be used to improve understanding of the causes of wildlife–vehicle collisions and result in models that reflect this understanding and recommendations that would reduce these collisions and wildlife roadkill in general. Priority. Create standardized electronic inventories of ex- isting crossings, bridges, and culverts and their geo-referenced locations in order to evaluate their potential for use by wildlife. Wildlife use crossings that are intended for them as well as transportation infrastructure such as culverts and bridge underpasses in order to avoid motor vehicles. Modelers, engineers, and biologists alike would be better able to distin- guish between the need for additional crossings or mitigation measures versus the modification of existing structures if there were state- and province-based electronic inventories of existing structures (culverts, underpasses) that could be ana- lyzed as part of a safety model for their potential and current use by wildlife. Priority. Develop methods to estimate the densities of animals near transportation corridors in order to calculate the risk of collision or exposure for certain stretches. The research team realizes that calculating “exposure” is a daunting task, and that surrogate measures, such as species density, daily movement behavior, seasonal migration patterns, annual har- vest records (see Mysterud177), and behavior near roads, would need to be linked with spatial landscape data to approximate “exposure.” This priority would entail working with state wildlife agencies in estimating and mapping where the most “high risk” animals are, i.e., deer, elk, and moose. Priority. Develop research guidelines on evaluating the effectiveness of wildlife crossings from a vehicle–animal col- lision perspective. The guidelines should demonstrate proper analysis methods and provide guidance on the monitoring of treatment sites. Monitoring of wildlife crossings should in- clude data on pre- and post-construction wildlife–vehicle collisions and roadkills. In order for models to evaluate the effectiveness of the full suite of wildlife crossings measures, monitoring efforts of these crossings measures need to be expanded. When crossings are installed, monitoring efforts have typically focused on documenting the number of ani- mals and species using the structures. The research team sug- gests that monitoring programs also include an analysis that documents pre- and post-construction wildlife–vehicle col-

127 lisions, roadkill carcass data, traffic volume, and possible wildlife exposure. Proper analysis methods need to account for numerous difficulties in analyzing collision data includ- ing regression-to-the-mean effects, spillover effects, differ- ences in crash investigation and reporting practice between jurisdictions when amalgamating data and exposure changes between before-after periods. Cost-effective designs for wildlife crossings need to be de- veloped through research and novel on-the-ground practices. An analysis of cost-effectiveness is a requirement for the consideration of most mitigation measures for wildlife. If flexi- ble standards or “standardized option-enabled” procedures and innovative designs could be created, there would be more opportunities to incorporate wildlife mitigation measures in transportation projects. The term “standardized option- enabled” means a general, clearly defined procedure or design with options so it can be modified to fit local situations. Currently it is difficult to link ecological values with safety values of wildlife mitigation measures for roads. Standardized proce- dures need to be developed for combining the estimated mon- etary costs of proposed wildlife crossings with ecological, safety, regulatory streamlining, and amortized monetary benefits. Standardized procedures would allow state and provincial departments and ministries of transportation to better evaluate how, what, and where to establish mitigation measures for wildlife in developing transportation programs and projects. Priority. Develop standardized procedures for estimating monetary costs and ecological, safety, regulatory streamlining, and amortized monetary benefits. Researchers in ecological fields need to work with economic researchers to better esti- mate the economic benefits of wildlife, intact ecosystems, and ecological processes. These values, once standardized in some manner, could then become part of cost-benefit analyses of mitigation measures. These analyses also need to include the amortized monetary benefits to society of reduced wildlife roadkill and vehicle collisions. These benefits would include reduced monetary costs to public agencies, insurance compa- nies, medical and personal costs to motorists, and increased wildlife populations available for recreational opportunities such as hunting and bird watching. Taking into consideration the economic benefit of including wildlife crossings early in project planning is necessary, for this approach can streamline environmental regulatory processes, thereby reducing overall project cost. Once these monetary benefits of mitigation measures are justified and standardized, a more realistic rep- resentative cost-benefit analysis method could be developed and employed across regions. Priority. Develop innovative and economically viable “option-enabled” alternative crossing designs after conducting standardized cost-effective procedures. Although there are dozens of wildlife crossings designs available, there are a stan- dard dozen or so in most common use. Through research and practice, option-enabled alternatives could be explored that may allow added permeability of the landscape over and under transportation corridors, while at the same time minimizing costs incurred. Ecological Considerations The genetic implications of the effects of roads on popu- lations are largely unknown, but theoretical and empirical evidence suggest that they fragment populations and their habitats. Transportation corridors are affecting the genetics of wildlife populations, the consequences of which are just beginning to be understood. There are a multitude of costs for wildlife associated with roads, from direct effects such as collision-caused mortality and habitat fragmentation to indirect effects such as decreased reproductive success and road avoidance. There are data that suggest the barrier effects of the roaded landscape reduce permeability of those areas for wildlife populations. Several studies have demon- strated that roads may act as barriers to small-mammal movements186,12,153 and as filter-barriers to large-mammal movements.23,105 Roads can be complete barriers to indi- viduals who cannot make their way across and whose road- related mortality can affect their small populations. This is especially true for populations of wide-ranging carnivores who are particularly vulnerable to road traffic collisions (Florida panther [Puma concolor coryi],155 ocelot [Leopardus pardalis],118 puma [Puma concolor],18 Iberian lynx [Lynx pardalis], 87 and wolves [Canis lupus]101). These effects over time will cause wildlife populations to suffer reduced sizes, isolation, skewed sex ratios (turtles222), depleted gene pools, and even extirpation. Indeed, concern has been raised regarding the influence of highways on normal mammalian distributional patterns and perhaps ultimately on speciation.10 For all that is known, there are still tremendous gaps in the understanding of just how the genetics of populations are being affected by the fragmenting and isolation effects of roads.197 The barrier effect of roads may reduce wildlife movement to the point of isolation, thereby reducing gene flow and increasing inbreeding and genetic drift.8,242 Current literature supports such theories that roads are causing genetic consequences for a variety of species. These species include wide-ranging grizzly bears (Urus artos)189,254 and black bears (Ursus americanus)256,231 and smaller localized species such as beetles,138 mice and shrews,170 voles,103 and frogs255. These and other studies indicate that research into the effects of genetic isolation due to transportation

128 corridors is necessary to begin understanding the conse- quences of roads and to mitigate their effects. Priority. Continue to study the genetic consequences of roads on wildlife. This research may most prudently be focused on wide-ranging species, small-movement species, isolated populations, carnivores, amphibians, reptiles, and small mammals. Directed research efforts into the restriction of movement and its genetic effects would help define the needs for freedom of movement for the target species. Elimi- nation of barriers to movement is essential for individual reproductive fitness and survivorship and has population con- sequences. Genetic research will help to define these move- ment needs, the necessary road crossing rates, and potential for appropriately designed wildlife crossings to help continue this flow. The research team argues that research will demon- strate that maintaining permeability of the landscape for a multitude of species will help negate the impacts of roads. There is a need for long- and short-term research tar- geted at assemblages of species to ascertain their reactions and behavioral adaptations over time to roads and associ- ated mitigation features. This research will inform the de- velopment of “option-enabled” general crossing designs that accommodate a wide range of species’ requirements. There is an urgent need for knowledge that would help in the design of wildlife crossings that allow the full range of wildlife species to move across and underneath transportation corridors. Information concerning behavioral reactions to roads and adaptations to crossings is lacking for most indi- vidual species, but particularly for species’ reactions in an associated community.58 For instance, even though mitiga- tion measures may be designed for specific wide-ranging and fragmentation-sensitive species (e.g., grizzly bear and lynx [Lynx canadensis]), there still are not sufficient design data to develop crossing structure guidelines for many of these species,79,56,58 much less suites of other species associated with the target species. Prior to developing guidelines for appro- priate mitigation measures, a better understanding of roads effects on suites of species is most desirable. To date there are relatively few studies of population-level and/or assemblage- level effects of roads. The existing studies suggest that the impacts can be significant. Findlay and Houlahan90 found sig- nificant effects of road density on species richness of wetland amphibians and reptiles, birds, and vascular plants. Fahrig et al.81 and Vos and Chardon238 found that presence/absence as well as density of local amphibian populations can be affected by road traffic. Forman et al.97 found decreased avian distri- bution and breeding near roadways in direct proportion to the volume of traffic on those roads. There are several groups of species for which there is a paucity of research and whose needs have not been ade- quately addressed. Work is limited for carnivores and small mammals (but see Clevenger and Waltho55,58), gallinaceous birds (turkey, pheasant, and grouse: for sage grouse see Lyon and Anderson151 and Connelly et al.60), and invertebrates (insects, spiders, worms) and for dispersing plants. Future options may be limited if the implications of roads on the survivorship of localized and low vagility species (e.g., mar- mots, bighorn sheep, and pikas) are not addressed. Gender responses to roads and crossings represent an unknown area of knowledge. These issues are only some of the many that need to be addressed. Another area of importance addresses the impacts of the noise of roadways and how it affects local and wide-ranging species, such as bears and neotropical migrants. Provocative work suggests that noise as indexed by volume and frequency has important negative effects on decreasing the bird species richness and diversity.199 The research team added a research component that addressed the road noise issue to the small- mammal research for this project. Priority. Continue research that addresses the reactions and adaptations of wildlife to roads and wildlife crossings. Research that examines the assemblages of species reactions to roads and crossings would be the most productive in relation to creating effective mitigation measures that allow the full range of wildlife species to move across and underneath transporta- tion corridors. Understanding the variables that contribute to wildlife behavioral reactions and how they may change over time is important. As transportation and natural resource pro- fessionals strive to create effective crossings that wildlife adapt to and actually use, consideration of extending monitoring efforts of crossings over several years in order to document the range of habituation and adaptation periods will be most beneficial. These efforts will be different among species and places. This priority can be addressed through specific regional wildlife–road research and also species-specific studies that may be broadened to include these objectives. Within each region of the country, the local scientists and wildlife and land managers are the professionals who can best address these questions, because wildlife reactions to roads and crossings vary from place to place. A crossing structure type that works for one popula- tion in a specific place may need to be modified to work effectively with another population and place. Regional research that addresses the effects of roads and their associated development, traffic, and noise on assemblages of species and those species reactions to mitigation measures would greatly contribute to creating effective mitigation measures that allow associated wildlife communities continued movement. The effects of roads and crossings on ecosystem relation- ships are largely unknown and need to be better assessed and understood. Road effects on ecosystems and landscapes need to be studied and quantified. Wildlife crossing mitigation measures

129 also need to be studied to assess their impact on ecosystems. Landscapes and ecosystems are affected by roads and other transportation structures synergistically with other human infrastructure, changed ecosystem processes, and changed wildlife and plant populations. The most obvious change to ecosystems is fragmentation. Fragmentation is a more diffi- cult phenomenon to evaluate than direct effects on specific species, and is analyzed over larger areas and greater time scales than most ecological studies. Forman et al.96,98 suggest using road density as a measure of fragmentation caused by roads. Road density is a simple spatial measure, providing an overview of the landscape.98 Other types of fragmentation measures could also be used to evaluate roaded landscapes. Further evaluation is also needed to understand how roads and mitigation measures influence and alter natural processes such as the flow of water, ecosystem dynamics (e.g., the rela- tionships between ungulates and their habitats), species interactions (e.g., predator-prey dynamics, see Little et al.149), population movement (e.g., movement to breeding areas), and individual behavior (e.g., the avoidance of roads by mothers with young, for grizzly bear see Proctor et al.189). Ecological effects are often indirect and multicausal and cannot be measured as easily as counting roadkill carcasses. This is in part the reason why relatively little is known about the effects of roads on ecosystem processes. Clearly, there is a need for a comprehensive synthesis that documents the indi- rect effects of roads on ecosystems and how these cumulative effects may in turn influence landscape permeability. Priority. Understand the effects of road density on the landscape for species of concern and ecosystems in general. For many species, roads generally reduce population sizes and increase the risk of population extinction. However, most species populations can persist in the presence of at least some roads. Therefore, in the context of road impacts on wildlife, probably the most important and most difficult question to answer is: what is the critical density of roads in an area below which a population of interest can not persist? This question is not easy to answer because of the spatial and temporal complexities of road impacts. As road density in- creases, wildlife habitat becomes increasingly fragmented.132 The numerical responses of large mammals to roads are gen- erally interpreted as responses to a road density threshold. Road densities above the threshold significantly reduce the probability for sustainable populations and coexistence. Several models have been developed to predict wolf pack oc- currence or survival in relation to road density in Minnesota and Wisconsin.230,166,171,172 A road density threshold of 0.45 km/km2 was identified that best classified pack and non-pack areas for wolves.171 Similar road density thresholds were reported for pumas and brown bears.237,54 However, these studies only scratch the surface of the problem of estimating critical road density. This is an area in which research is ur- gently needed. Other ecosystem components affected by roads could also be measured with road density, including peak flows in mountain streams98, erosion, and the spread of invasive plants and the subsequent impacts for ecosystem in- tegrity, to name a few. Road density is a simple measure, but road impacts on ecosystems vary considerably with traffic volume, speed, and infrastructure width, surface, and design.129 For example, Foreman and et al.97 found grassland birds avoided regular breeding in patch edges near roads in direct proportion to road volume, moving breeding activities farther away (up to 1 km away) from roads with greater vehicle numbers per day. In order to gain a more thorough understanding of such road effects, the research team suggests examining the properties of the roads in conjunction with density to ascertain the eco- logical relevance of each road.115 Other aspects of the roaded landscape could be analyzed for impacts to ecosystem function. Analyzing the specific form or spatial pattern of the network of remaining natural patches and roads could reveal ecosystem properties.98 This could be accomplished in part through the use of indices of patch size or mesh size.132 Different mesh/patch sizes of natural areas contribute to different ecological conditions.98 With such in- dices, studies could be compared and contrasted to evaluate how roads are affecting ecosystem function and the basic eco- logical processes such as water flow, disturbance regimes, predator-prey interactions, seed dispersal, and movement among populations. These effects could be summed over ecosystems to find the cumulative costs of roads over regions. Priority. Measure the effects of wildlife mitigation meas- ures on ecosystem dynamics. These assessments could be performed by monitoring specifically chosen ecological indicators at different levels of biological organization: genes, individuals, populations, and species across landscapes. Assessments would be performed both before and after place- ment in order to judge the effectiveness of actions aimed at connecting communities and populations and possible cumu- lative effects.58 Ecosystem assessments using specific ecological indicators would benefit from standardization and accuracy testing in order to obtain a tangible conservation value of the studied crossing structure. That would allow assessment of the ecological value of mitigation measures and possible cost- benefit values of potential crossings. Examples of questions to be answered include but are not limited to: How does the wildlife crossing structure affect the predator-prey dynamics within an ecosystem? Does the presence of artificially increased vegetative cover near passages change the use of these areas by cover-associated species? How does placement of a crossing structure influence the willingness of target species to use it? For instance, if a structure is placed along a riparian area does

130 it promote the passage of some species and individuals while hampering the movements of others? The accomplishment of the priorities in this task will take the concerted effort of many scientists. The larger-scale landscape context of road effects and transportation programs needs to be addressed through con- nectivity analyses at the state/provincial and regional levels. There is a need for all states and provinces to conduct state/province-wide connectivity analyses to help determine “fracture zones” among conservation areas that can then be prioritized in transportation programs for mitigation efforts. These fracture zones are where transportation corridors bisect natural wildlife movement corridors and potentially restrict movement and permeability of the natural world. There appear to be few large-scale state- or province-wide landscape approach efforts to address the effects of the roaded landscape on wildlife and ecosystems. Although the concept of “context-sensitive planning” is gaining national attention within the transportation community, it does not appear to the research team to explicitly include the surrounding wildlife habitat. The research team believes connectivity analyses create a window of opportunity to include ecosystem-level and landscape-scale considerations in transportation programs and individual projects. Without a proactive approach, future measures aimed at patchwork retrofitting and restoration may remain a poor second choice to properly planned and maintained landscape permeability in most regions. Currently, landscape-scale connectivity analyses have been conducted in a variety of formats in several states and provinces, including California,187 Washington,212 Montana,205 Colorado, Utah, Arizona, New Mexico, and eight southeastern states.46 These analyses involved landscape linkage models and the creation of GIS-generated maps, or workshops aimed at addressing statewide connectivity, or rapid assessment work- shops centered on specific roads, where professionals from across the state met to identify and map all potential major landscape linkages within the state and the roadways that po- tentially fracture these connections. The research team suggests that similar efforts be conducted for all states and provinces and the results incorporated into spatially explicit statewide databases and programs so these maps and accompanying data can be used in DOT and MoT planning for linking mitigation and implementing changes in their long-range programs.80 Priority. Researchers, agency personnel, non-profit organizations, and the public together can create and dis- seminate state- and province-wide connectivity analyses. The research team suggests collaboration in conducting the science, widely attended workshops to enhance needed information exchange, and partnerships to fund these efforts. A likely partnership for connectivity analyses fund- ing could include states DOTs and provincial MoTs and their wildlife agency counterparts who can benefit greatly from such analyses. This type of effort has worked in several states. Regional approaches may work best. An effective type of analysis might include GIS models that analyze landscape linkages based on four important variables: focal species movement patterns, land cover, human density, and road density,212 or a more inclusive list of environmental vari- ables.215 Digital topographic data can also help identify movement corridors in places containing drainages and ridgelines. Finally, the collective knowledge of land man- agers, wildlife biologists, non-profit environmental organi- zations, and state DOT/provincial MoT professionals can be brought together in critical connectivity workshops where the participants can work synergistically to identify key land- scape linkages and the transportation corridors that frag- ment them, and prioritize projects needed to restore wildlife and ecosystem permeability. In light of the amount of progress that has been made in these workshops in the past 2 years,69,204,203,205 this priority holds great promise. A continent-wide set of guidelines is needed for defining specifics in the consideration, placement, design, mainte- nance, and monitoring of crossings and other mitigation measures. There is a need for research to aid in the development of guidelines to facilitate the planning, placement, design, maintenance, and monitoring of wildlife crossings across North America. Transportation planners, engineers, and bi- ologists need guides to effectively mitigate for the effects of roads on all wildlife species within affected communities. Although wildlife crossings have been built for more than three decades, there is no standardized set of guidelines to assist these professionals and other agency personnel to mit- igate for terrestrial and aquatic wildlife.11 A North American set of guidelines for wildlife crossings would include specifics on conditions that trigger the consideration of mit- igation measures; how and when to plan for structures; where to locate mitigation measures; design considerations; how to combine several types of efforts such as fences, underpasses, and ROW escape structures; standards for monitoring and maintaining structures; and how to meas- ure the success of projects. Priority. Define the necessary conditions for considering when to identify areas in need of wildlife crossing mitigation measures. Predictive models or threshold requirements would help determine when (a) crossing structure(s) is (are) needed to help mitigate for certain volumes of traffic, safety consider- ations, roadkill hotspots, the presence of endangered, threat- ened species or species of special concern, landscape linkages fracture zones created by transportation corridors, and

131 the presence and need for movement of surrounding wildlife populations throughout critical habitat. Priority. Engage the research community in the devel- opment of guidelines for the placement of crossings. Scien- tists and wildlife managers and biologists need to critically review the habitat-based linkage or movement models and rapid assessment techniques currently used to identify passage placement, and identify a suite of possible methods for practitioners. Emphasis should be on criteria to locate mitigation measures.111 Priority. Design considerations need to be adequately ad- dressed for the full suite of crossings. There is a need for research to help in the selection of target species, and the de- termination of the number, size, and dimensional character- istics of structures needed within an area to help maintain maximum permeability for the suite of associated species.80 Design guidelines for mitigation measures associated with crossings are also needed. Considerations include determin- ing the required lengths of fences erected to guide wildlife (both large and small) to crossings; addressing the suitability of establishing or eliminating median islands in conjunction with crossing areas; creating underpasses with a naturally lit open space in the median of divided highways, in effect creat- ing two underpasses under travel lanes rather than one long darker underpass; taking into account other nearby trans- portation corridors such as railways; retrofitting existing culverts for fish and other aquatic species; and possible alternatives or complements to crossings such as remotely sensed, active lighted warning signs, possible crosswalks over low-volume roads, the clearing of vegetation, temporary clo- sure of roads, public transit options, reduced-speed zones, and the elimination of certain roads (road decommission). Indeed several measures may be coupled for maximum effectiveness. Priority. Monitoring standards for crossings need to be researched and created. Bank et al.11 suggest a national U.S. policy requiring post-construction monitoring and mainte- nance measures for wildlife. Most existing structures have seldom if ever been monitored, or have been only sporadi- cally checked to determine if they have served their purpose.111 Guidance on monitoring efforts and temporal specifications would greatly assist managers, planners, and biologists and allow for comparable analyses among struc- tures to ascertain their efficacy. The majority of past and current monitoring projects have been conducted in concert with academic institutions and the U.S. Fish and Wildlife Service. This monitoring, if done correctly, is essentially research. Future monitoring research could be standardized, implemented by, and mandated for future projects by the state wildlife and transportation agencies and federal agencies, including resource-based agencies such as the U.S. Fish and Wildlife Service, Environment Canada, Fisheries and Oceans Canada, and transportation entities, for example the FHWA and Transport Canada. The overall standardiza- tion of monitoring projects would mean a commitment of necessary funds from U.S. Federal Highways Program and other sources. The expected benefit would be an enhanced understanding of which structures work most effectively in specific situations. Priority. North American guidelines for crossings need to include methods for defining success and effectiveness.11 Defining success would involve addressing the number of individuals (including the difference between males and females and juveniles and adults) of a target species who have used a structure, number of species found to use a structure, use by endangered and other high-needs species, reduction of wildlife–vehicle collisions, as well as other measures. Fish pas- sages created in retrofitted and replaced culverts and bridges along streams have been evaluated through a quantifiable checklist of goals accomplished: e.g., the number of a certain species using the new passage, the number of kilometers those species have traveled upstream, how many individuals breed and re-populate a specified river distance within a watershed. These kinds of quantifiable measures present an objective method for assessing wildlife–landscape permeability across roadways and would greatly improve the credibility of wildlife crossings science and practice.