As noted above, this literature survey reviews material produced since 2003 and looks specifically at new research conducted on data analysis procedures and mitigation measures.
A considerable amount of literature is produced annually on this topic. Much of this literature is presented at annual and bi-annual conferences and events sometimes dedicated solely to this issue including:
- International Conference on Ecology and Transportation (ICOET) – Bi-Annual events http://www.icoet.net/
- Deer Vehicle Crash Information Clearinghouse – Annual workshop
- Transportation Research Board Annual Meetings – Workshops and poster sessions
Some University Research Centers specialize in the issue, creating the impetus for more research. These include:
- Center for Transportation and the Environment http://itre.ncsu.edu/cte/index.asp
- University of Wisconsin http://www.deercrash.com/index.htm
- Western Transportation Institute – Montana State University http://www.coe.montana.edu/wti/
In 2005, Montana State University developed a Road Ecology Curriculum and implemented a university-based road ecology course.
The literature indicates a fair amount of research continuing to explore new mitigation measures. The need to demonstrate the potential for new technologies to address the problem creates the catalyst for some of the research conducted since 2003. The research is also extending far beyond the motor-vehicle-large-animal collisions with expansion to include animal/roadway interaction in general. The expression Road Ecology seems to be accepted as the appropriate terminology encompassing this field of activities.
The underestimation of the number of collisions between motor vehicles and large animals, especially deer, seems to be increasingly well accepted even though no firm statistics exist to validate the statement and not much research is dedicated to addressing this concern.
From the considerable volume of literature on the subject the following are highlighted as they are substantive and/or bring a new research perspective. The entries are made in chronological order.
New research related to data collection, analysis and modeling techniques includes:
National Cooperative Highway Research Program – NCHRP (Active Project) launched a new project on the animal-vehicle crash issue. This new project is a synthesis examining how animal-vehicle crash data are collected and used in the United States and Canada The synthesis will survey state and provincial agencies about their collection and use of data. This will include but not be limited to the following items:
- What kinds of AVC data (crash, collision, near miss, carcass removal, etc.) are collected?
- Are domestic animals included?
- How are the data collected?
- Who collects the data?
- Where are these data stored and managed?
- How are the data accessed, reported and used?
- Are these data shared or combined among agencies? Who are they?
- Are system improvements being discussed and/or planned? What are they?
- What are the bureaucratic, political or procedural obstacles to and opportunities for implementing, advancing or improving data collection?
- Hasson (2005) of the Federal Highway Administration developed a national perspective on the issue of highway safety and wildlife. In his presentation he presented the national statistics and stated that animal-vehicle crashes are probably under-reported by 50%. He confirmed that these crashes are a growing problem and that the data needs to improve. The presentation also outlined the role of the U.S. federal government on this issue and the possible animal-vehicle crash-related activities included in the new legislative agenda. Above all, the presentation stressed the need for better data and communication of best practices.
Knapp (2005) presented a paper at the 2005 Transportation Research Board Annual Meeting: Defining the Deer-Vehicle Crash Problem in the United States: National Estimates and Regional Data Collection. The paper presents data from Illinois, Iowa, Michigan, Minnesota, and Wisconsin since 2001, and concludes that deer-motor vehicle collisions are an increasing transportation safety concern throughout most of the United States. The actual magnitude of this problem, however, can only be roughly estimated. Knapp recommends the magnitude, trends and/or location of collisions should be properly documented in each state, regionally and/or nationally and gaps in the datasets identified. The diversity and multidisciplinary nature of the related data that are available require close communication and coordination among multiple agencies, researchers and interested parties.
The paper examines data for the five, above-mentioned states and notes that the total number of actual deer-vehicle crashes may be at least twice as large as reported. In Minnesota, it is believed to be three to four times as large as reported. The number of unreported deer-vehicle crashes probably varies from state to state due to different reporting procedures with few states tracking the actual number of carcass collections.
- Wells (2005) of the Insurance Institute for Highway Safety presented a paper at the annual workshop of the Deer Vehicle Crash Information Clearinghouse on the characteristics of vehicle-animal crashes in which vehicle occupants are killed. The paper shows that, in the United States, from 1993 to 2003, the annual number of fatalities resulting from these collisions increased from 95 to 195, a 100% increase. The author calls for the development of a strategic agenda to address this issue.
- Hardy (2005) published an analysis of wildlife-vehicle collision data related to applications for guiding decision-making for wildlife crossing mitigation and motorist safety. Research on wildlife-vehicle collisions has shown that they do not occur randomly but are spatially clustered. The report showed that the presence of wildlife tends to be linked to specific habitats and adjacent land-use types. Thus, landscape spatial patterns would be expected to play an important role in determining road-kill locations and rates. The project used wildlife-vehicle collision data to demonstrate how this information can be used to aid transportation management decision-making and identification of appropriate mitigation measures.
- Huijser (2005) focuses on developing a national standard for recording data relative to animal-vehicle collisions using a hand-held device combined with a GPS system. The project’s objective is to develop a national standard for the reporting of animal vehicle collisions thereby encouraging transport departments and others to collect these data and to provide better integration and analyses of the data. It is believed that more accurate and consistent data will also help transportation and wildlife managers to prioritize and focus efforts to reduce collisions.
- The California Department of Transportation (2005) has recently contracted Western Transportation Institute to develop and implement an at-scene data collection and incident support system to transmit data, including photos, from incidents.
Sielicki (2004) presented 20 years of data using BC Wildlife Accident and Reporting Systems (WARS). The WARS system is designed to analyze wildlife accident data collected by maintenance contractors on numbered highways in British Columbia.
The WARS system is becoming an increasingly valuable information resource for BC Ministry of Transportation, other government agencies, consultants, researchers, wildlife associations, special interest groups and members of the general public. The Ministry of Water, Land and Air Protection uses WARS data to assess provincial wildlife population trends. The Insurance Corporation of British Columbia (ICBC) uses WARS data for identifying highway locations where joint BC Ministry of Transportation and ICBC initiatives, such as exclusion fencing, warning reflectors, and infrared camera detection systems, can be targeted to reduce wildlife-related motor vehicle collisions. The success of the WARS system in British Columbia has made it a model for other agencies seeking to monitor wildlife-related motor vehicle accidents.
Knapp (2003) completed a survey on behalf of the Deer-Vehicle Crash Information Clearinghouse. The survey objective was to investigate Department of Transportation and Department of Natural Resources activities related to the collection and management of deer-vehicle crash and related data.
Law enforcement personnel complete the majority of the crash reports in each state, but all five of the states also allow some form of citizen self-reporting of a crash. The survey results confirm that only two states do not include this self-reported information in their official crash database. The status of this information in the other three states is unknown. The validity of crash information reported by the general public (away from the incident scene), however, should be questioned. Four of the states in the region allow a crash to be identified as a DVC (Deer-Vehicle Crash) on their crash report. The fifth allows animalvehicle crashes to be specified.
Newhouse (2003) presents the efforts deployed in British Columbia using infrared technology as a mitigation measure. In the publication, the author claims that in British Columbia, about 16,000 animal-vehicle collisions occur annually, including unreported cases.
The Wildlife Protection System (WPS) uses infrared cameras to detect wildlife on or near highways. When wildlife is detected, flashing lights are triggered, warning drivers to reduce speed and anticipate wildlife on the roadway. The first trial was initiated in the summer of 2002 in Kootenay National Park, British Columbia, Canada. In the trial, a camera was mounted on a 6-m pole at each end of a 2-km stretch of highway. Adjacent to each pole was a trailer containing a computer (with tracking software), two radar guns, and a conventional digital video camera. Continuous (24-hour) infrared and conventional video footage was recorded. In addition, an “event log” was generated in an Excel spreadsheet that recorded traffic speeds before and within the test zone, and animal detections within the zone.
- Elzohairy (2003) in his paper presented at the Transportation Research Board Annual Meeting on the Characteristics of Motor Vehicle-Wild Animal Collisions in Ontario states that despite reductions in the total number of traffic fatalities and in the total number of reportable collisions in Ontario, there has been an increasing trend in motor vehicle-wild animal collisions. The number of collisions increased by 50 percent over a 6-year period (1996-2001) and claimed 42 lives over that period. The paper presents an analysis on collision trends and patterns associated with motor vehicle-wild animal collisions. It also reviews approaches and techniques currently available to reduce the risk of fatality and injury, as well as offering recommendations for future research.
The following notes efforts that have also gone into the modeling of the events.
- Knapp (2005) looked at the characteristics of the data and advanced the need of a Countywide Deer-Vehicle Crash Frequency Model using a negative binomial regression approach. The frequency model developed showed an increase in crash with deer population and vehicle travel, and a decrease with increased estimates of wolf population and woodland acreage. Deer population and vehicle travel approximate crash exposure measures, but wolf population and woodland acreage were also significant and added strength to the model. The authors claimed that the modeling approach used is more valid for crash data than those used in the past, and the model developed predicts a generally accepted measure of safety.
- Meyer (2004) looked at roadside characteristics and the roadway. Meyer’s methodology compared the relative risk of road segments so that they can be prioritized. Forty-five predictor variables were considered, and the predicted variable was accidents per year per mile. Wooded-land area by the side of the roadway, number of lanes, median types, traffic volume, posted speed, clear width, number of bridges and/or visible culverts, roadside adjacent side slope, roadside topography in the transverse direction, presence of deer warning sign and traditional fencing were all found to be positively correlated.
- Kline (2003) in a paper presented at the 2003 ICOET conference describes a model for estimating wildlife mortality on roads. In this instance, the research was undertaken to better understand the effects that roads are having on wildlife in Saguaro National Park. Surveys were conducted from 1994-1999. A model to estimate the average annual number of animals killed on roads in and adjacent to the park was developed.
- Gunson (2003) published a report showing the patterns and characteristics of large animal-vehicle collisions in the central Canadian Rocky Mountains. The paper focused on understanding the patterns and processes that result from animal-vehicle collisions.
Other efforts have gone into developing databases and clearinghouse information on the same issue:
- Western Transportation Institute at Montana State University (2003) initiated a project referred to as ARTEMIS. The project’s objective is to develop a database that would allow other universities, transportation professionals and interested individuals access to a complete reference source focused on animal-vehicle collisions and mitigation options. To date, not much has transpired from this initiative.
- Knapp (2003) published a summary of the activities of the Deer-Vehicle Crash (DVC) Information Clearinghouse. The Clearinghouse was initiated in July 2001 by the Wisconsin Department of Transportation as a regional information center. Five states in the Upper Midwest (i.e., Michigan, Minnesota, Illinois, Iowa and Wisconsin) joined in with this project. One of the first projects of the DVC Clearinghouse was to summarize the current state of the knowledge related to deer-vehicle crash countermeasure effectiveness.
The literature also contains information on mitigation measures and general improvements to the state of knowledge on the animal-vehicle collisions issue:
- Huijser (Active Project) is working on a quantitative comparison of different types of animal detection systems with regard to system reliability, and operation and maintenance aspects.
- National Cooperative Highway Research Program – NCHRP (Active Project) This project is an evaluation of the use and effectiveness of wildlife crossings. The objective of the project is to develop guidelines for the selection (type), configuration, location, monitoring, evaluation and maintenance of wildlife crossings providing transportation professionals a better understanding of how wildlife and fisheries issues can be integrated into planning, engineering, design and maintenance of highways.
- Clevenger (2005) produced new guidelines for wildlife crossing structures. The new guidelines are intended to assist agencies in incorporating wildlife fencing and crossing systems into highway and road designs thereby reducing the harmful impacts of transportation facilities on wildlife and reducing collisions. The project is funded by the Federal Highway Administration.
- Huijser (Active) worked on a project to deploy and evaluate effectiveness of automated animal detection and warning systems. Western Transportation Institute will evaluate different types of animal detection systems from different vendors at the same site and under similar circumstances. Western Transportation Institute will also evaluate a relatively new mitigation measure called “Roadside Animal Detection Systems” (RADS).
- Hardy (Active) conducted research on whether or not fences and cattle guards are effective at reducing the number of animal-vehicle collisions, and at re-directing animal movement patterns through existing highway ‘crossing’ structures (e.g., road and railroad bridges and culverts). The results of this project are expected early in 2006.
- Hardy (Active) evaluated the effectiveness of Intelligent Transportation Systems, combined with public information campaigns, to increase awareness of the high risk of animal-vehicle collisions on the Bozeman Pass and reduce the number incidents. The ITS technology will make use of Dynamic Message Signs (DMS) to motivate drivers to decrease speed and be aware of animal crossings. The results of this project are expected early in 2006.
Knapp (2005) presented a paper on crash reduction factors for deer-vehicle crash countermeasures at the Transportation Research Board Annual Meeting and suggested safety research needs. The paper suggests an approach to deal with existing mitigation measures:
- Countermeasures Used with Conflicting Safety Analysis Results and “Tried”: The paper recommends that a properly funded, designed and documented evaluation of these countermeasures (i.e., deer whistles and roadside reflectors/mirrors) be completed to definitively determine and quantify what, if any, collision reduction effectiveness they may have.
- Countermeasures Used with Generally Positive Safety Analysis Results and “Proven”: The paper recommends that the impacts of exclusionary fencing/wildlife crossing installations continue to be evaluated.
- Countermeasures Used but Rarely Studied for Safety Impacts and “Tried”: the past safety evaluations of these countermeasures has been limited in their approach and number. Additional evaluations are needed to determine their actual impact on collisions. Replicating and improving upon the studies previously completed to refute or support their results is necessary.
- Countermeasures Used but Rarely Studied for Safety Impacts and “Experimental”: New deer crossing sign designs and sign/technology combinations only used in pilot study situations. The approach used in their safety evaluations should be reviewed to ensure they use the most appropriate analysis methodologies.
- Countermeasures Used but Not Studied for Safety Impacts and “Tried”: countermeasures that have not had their impacts studied. These measures include public information/education and roadway development (i.e., maintenance, design and planning) policies.
- Countermeasures Used but Not Studied for Safety Impacts and “Experimental”: Two countermeasures are also being used but are considered more of an “experimental” rather than a “tried” safety strategy. Their impacts on collisions have not been studied, but they are still believed to be at the “pilot study” stage of implementation. These measures include in-vehicle technologies and deicing salt alternatives.
- Countermeasures Not Generally Used but Rarely Studied for Safety Impacts and “Experimental”: Four countermeasures have been suggested for implementation, but are not generally used. Their impacts have only rarely been studied.
Huijser (2004) provides an overview of animal detection and animal warning systems in North America and Europe. In the United States animal-vehicle collisions are estimated to cause 211 human fatalities, 29,000 human injuries and over one billion dollars in property damage a year. Similar numbers are available from Europe (excluding Russia) where the annual number of collisions with ungulates was estimated at 507,000. These collisions were estimated to cause 300 human fatalities, 30,000 human injuries and over one billion dollars in material damage per year. The report identified 27 locations with an animal detection or animal warning system. Nine of these sites are located in North America and eighteen in Europe.
This overview shows that a wide variety of animal detection and animal warning systems have been installed across North America and Europe. Many of the systems encountered technical problems or experienced false positives, false negatives or maintenance issues. The report maintained that this was to be expected since most animal detection and animal warning systems are new applications of relatively new technology. In addition, the systems are typically exposed to rain, snow, heat and frost. A few systems seem to have resolved most of the problems and operate well. Examples are systems in Switzerland and in Finland. In Switzerland, passive infrared detection systems were able to reduce the number of animal-vehicle collisions by 82 percent.
The author concludes that it is important that animal detection systems produce very few false positives and false negatives. False positives may cause drivers to eventually ignore activated signs, and false negatives present drivers with a hazardous situation. Driver response through reduced vehicle speed or increased alertness determines how effective animal detection systems really are.
The report goes on to say that minor reductions in vehicle speed are important since a small decrease in vehicle speed is associated with a disproportionately large decrease in the risk of a fatal accident. In addition, activated warning signs are likely to make drivers more alert. Driver reaction time to an unusual and unexpected event can be reduced from 1.5 seconds to 0.7 seconds if drivers are warned.
The report concludes that animal detection and animal warning systems have the potential to be an effective mitigation tool. The report notes that each system type has its own (potential) strengths and weaknesses, and one has to review them carefully before installing a system in a particular location. In addition, further research and development is needed before animal detection and animal warning systems can be applied on a wide scale.
- Rogers (2003) examines using a GIS system to better understand the spatial and temporal patterns of deer-vehicle accidents within a town and to create a deer-vehicle accident management plan for a town. The results of the research showed that GIS analysis was able to show that measures being deployed had a significant effect.
- Dodd (2003) provides an evaluation of measures to minimize wildlife-vehicle collisions and maintain wildlife movement across highways in Arizona. The research objectives were to determine the effectiveness of the full complement of measures to minimize the incidence of wildlife-vehicle collisions and evaluate the degree to which wildlife permeability across the highway is maintained. The results of the research were to provide ongoing construction implementation guidance to Arizona Department of Transportation project managers throughout all construction phases of highways.
- Van-Riper (2003) presented a paper on Animal-Vehicle collisions in Maine at ICOET 2003. The paper showed that while the number of all other types of crashes is dropping, those with large wildlife species are increasing. It also stated that this total is probably a low estimate of the actual number of crashes since data used were from official accident records only. The paper also points out the fact that mitigation methods have had limited success, and there was an overall lack of statistically rigorous monitoring to evaluate the efficiency of the method(s). The report concluded that no simple solutions were apparent.
- Gordon’s (2003) paper on motorist response to deer-sensing warning systems in Wyoming showed exactly the difficulties related to new technologies. It found that while the sign may work well for local traffic, people passing through the crossing will encounter the sign only once and during their brief encounter probably will not fully understand how it functions. Additionally, a program educating the local citizens about how the system works is recommended in conjunction with the system's installation.
The literature shows continuous attention being given to the issue of animal-vehicle collisions in North America. New research projects initiated by some high profile agencies such as NCHRP and the US Federal Highway Administration demonstrate the level of interest the issue is gathering in the United Sates. The literature also showed that the issue of underestimation seems to be recognized and acknowledged. This is also supported by an increased awareness regarding the need for data quality. The document published by FHWA recognizes this.
Along the same lines, new research points to the need for improved methods for recording animal-vehicle collisions. The Wildlife Accident Reporting System (WARS) data collection method developed by British Columbia Transport Department is often mentioned as a model for data collection. The literature notes that data collection is often done by two and sometimes three different organizations within the states and provinces. The use of a more common approach and/or technology is now being looked at in the United States.
The literature also shows that there has been continued development and testing of mitigation measures with some success with detection technologies (e.g. Switzerland site with an 83% reduction). It also shows that although much progress has been made in the identification of more universal mitigation measures, local conditions often demand custom-designed solutions