Best Practices Manual: Wildlife Vehicle Collision Reduction Study

CHAPTER 2: STATEWIDE/REGIONAL PLANNING APPROACH AND DATA NEEDS

Efforts to substantially reduce WVCs would be best built on a foundation of statewide or regional planning (figure 2). This chapter provides a broad overview of planning efforts and data needs aimed at WVC reduction. Conducting statewide/regional data collection programs (section 2.2) and identifying regional priority areas (section 2.3) are discussed in detail.

Graphic image depicting sections Ch 2.1-2.4 as explained in the paragraph above.
Figure 2. Regional planning as part of a strategy for reducing WVCs.

2.1 REGIONAL PLANNING: A PROGRAMMATIC APPROACH

A Statewide and regional planning approach to achieving WVC reductions should ultimately result in implementing mitigations discussed in chapters 4 and 5.Projects implementing specific mitigations at specific locations should be prioritized and implemented as funding is available.

A more opportunistic approach could include an annual check of the statewide transportation improvement program for reconstruction projects that cross WVC regional priority areas. Incorporating WVC mitigations into a planned reconstruction project will be much less expensive than installing them as separate projects.

A checklist for a statewide WVC reduction program can be found in chapter 8.

One suggested approach to ensure that the WVC reduction program is integrated into long range planning is to establish a state or regional multi-agency committee at the local Metropolitan Planning Organizations, regional planning offices, and/or State DOTs. These committees should have experts in this area providing input and oversight. Such an oversight committee should establish goals for the WVC reduction program, gather the needed data and maps to establish a baseline assessment of the magnitude and nature of WVCs to support these goals, and develop a strategic approach to integrating these goals into transportation planning Successfully integrating WVC reduction into planning not only requires the appropriate data but also a data analysis procedure.

2.2 STEP 1: COLLECT STATEWIDE/REGIONAL DATA

While data collection and monitoring do not have direct benefits in reducing WVCs; good data on the magnitude, trend, location, and type of WVCs clearly highlight the issue, document the need for mitigations discussed in this manual, and allow for a procedure to prioritize possible mitigation locations. It is the best way to ensure that appropriate mitigations will be installed at the locations where they will have the most impact. WVC data are crucial in justifying and prioritizing locations for mitigation. Additionally, in order to utilize the most effective mitigations for WVCs, the effectiveness of the measures must be evaluated (described further in chapter 6).

Where to Find the Crash Data

There are three primary sources of WVC data: insurance data, animal carcass counts, and crash reports. Insurance data can give an idea of the magnitude of the problem, but is not spatially referenced and rarely is used on a local level.

A good example of a carcass data collection program is the Maryland Department of Transportation's Large Animal Removal Reporting System LARRS program. This program is implemented administratively from the top down, so there is a consistent standard across the department. Questions regarding carcass collection are part of a broader form that maintenance workers are required to complete for general maintenance tasks, increasing the likelihood of capturing the data. For information on the Maryland LARRS carcass data collection method, contact William Branch, Maryland Department of Transportation, (410) 545-8626, WBranch@sha.state.md.us.

To assist with carcass data collection, handheld computers (personal digital assistants) equipped with global positioning system (GPS) software have been developed, mostly as prototypes. One such system is the Roadkill Observation Collection System (ROCS), currently under development by the Western Transportation Institute (WTI). For more information on handheld GPS units for automatically collecting data (ROCS), contact Rob Ament at WTI, (406) 994-6423, rament@coe.montana.edu.

Crash data from police crash reports and statewide crash databases have the benefit of being readily accessible and spatially referenced. Because of standardized reporting and recording methods, this data source is fairly uniform within a state.

Crash Data Issues

The NCHRP Project 20-05-Topic 37-12 provides a current state of the practice on WVC data collection. ⁵ A survey conducted as part of this report asked State departments of transportation and departments of natural resources if they collected WVC data (crashes or carcasses). Of the 30 States that responded to the survey, 19 collected crash data on WVCs, 13 collected carcass data, and eight collected neither. It should be noted that although the States reported they did not collect data, a review of crash reports shows that every State except one has "animal" as a check box on the crash form, as mentioned above. Of the 30 responding departments of natural resources, nine said they collected crash data, 15 said they collected carcass data, and 12 said they collected neither. Responses indicated there are substantially less data collected on local and collector roads compared to interstate highways and arterials. Many States discussed problems with the data, including inconsistencies, location accuracy, and underreporting.

Huijser and others found that every State but one had the word "animal" on their crash reports, allowing those reports to be separated from reports involving other types of crashes. ⁵ However, the quality of the data needs to be considered. Inconsistencies in the manner of data collection among the States can affect the reliability and precision of the data. For instance, many crash reports include domestic animals in this category and, thus, may not be a true indication of WVCs. Another example is that "animal" may be simply a check box under entries asking the investigator to indicate the "most harmful event," "contributing factor" or "first/second/third object hit. " If "animal" is under the "most harmful event" category, a crash where a driver swerves to miss an animal and collides with a fixed object on the side of the road may not be recorded as an animal crash.

The largest challenge with using reported crash data is that it under-represents the actual number of crashes. Nationally, approximately 300,000 WVCs are reported through crash databases each year. Extrapolating from carcass counts and insurance industry numbers, it is more likely that between one and two million WVCs occur annually. Because of the relatively small sample sizes with crash data, some States utilize carcass data to allow better identification and prioritization of problem road sections.

In addition to wildlife road mortality data, wildlife movement data can also help guide WVC reduction efforts and has the added benefit of providing insight into conservation concerns.

2.3 STEP 2: IDENTIFY AND PRIORITIZE REGIONAL/STATE ROAD MORTALITY AND CROSSING LOCATIONS

The focus of this section is identifying hotspots (or priority locations) on a statewide or regional scale. Identifying priority locations on a corridor scale is discussed in section 3.4.Regional planning for WVCs involves a programmatic approach to prioritizing problem areas in order to maximize investment in mitigations. Many WVC reduction efforts focus on identifying and prioritizing WVC hotspots within a region or State. Including both WVC hotspots and habitat linkage zones in planning efforts provides better insight into animal movements, resulting in both safety and conservation improvements.

WVC hotspots are locations with high incidence of WVCs. A hotspot would be identified as a segment of road meeting some threshold value for the frequency of WVCs. Addressing WVC hotspots have both a human safety (fewer and less severe collisions) and conservation component (fewer animals killed or injured).
Habitat linkage zones are zones that link areas with core habitat for selected species. Depending on the species, the number and size of the habitat patches, and the connectivity between these patches, a certain minimum level of wildlife movements between the individual habitat patches can be vital for the long term survival probability of fragmented populations (meta-populations).

Habitat linkage zones may include but are not necessarily restricted to WVC hotspots as WVC hotpots ignore locations where animals may be crossing the road successfully and where animals may want to cross, but shy away from the road corridor once they get close (aborted crossing attempt).

Identifying WVC Hotspots

Available carcass data provide a good indication of the occurrence of WVCs if the search and reporting effort for the road section concerned is constant. Keep in mind that consistency and detail of the data may vary among the individuals recording the data and across maintenance districts (i. e. , some maintenance personnel or districts may be more diligent in recording carcasses).

Currently, state safety management systems collect and analyze crash data to identify safety hotspots typically utilizing an exposure rate such as crashes per million vehicle-miles travelled. Using a traffic exposure rate may not be appropriate for targeting WVCs because, following the underlying logic of using an exposure rate, the number of animals that cross the road, typically an unknown factor, should be considered along with the volume of traffic. Because of this and the fact that WVCs occur more often on two-lane, low-volume, rural roadways, a straightforward frequency per kilometer (or mile) per year may be a good measure to identify hotspots.

Identifying and Prioritizing Habitat Linkage Zones

Some States have identified and prioritized these habitat linkage zones (e. g. , developed wildlife habitat linkage plans). Based on the local situation, linkage zones can be recognized, and proper planning put in place to protect and strengthen these linkage zones. Alternatively, if the natural habitat has almost or completely disappeared (e. g. , because of large scale agriculture), then one may consider re-creating (semi-)natural habitat and habitat linkage zones which may direct wildlife movements to specific pre-determined locations where mitigation measures have been provided. By prioritizing these linkage zones, mitigations can be focused on the most critical areas. Mitigations may be aimed at WVC reduction, improved habitat connectivity or, ideally, both. Improving or restoring habitat linkage zones may include mitigation for roads and traffic, but other factors that affect the presence and quality of a habitat linkage zone may need to be addressed as well before a habitat linkage zone may function at the desired level. Such factors may include human presence and disturbance because of agriculture, (sub)urbanization, and other types of habitat alteration or direct or indirect disturbance. Depending on whether an integral approach is taken, addressing such other factors may or may not be part of a transportation project. Note that habitat linkage zones may relate to large as well as small species, including species that may not be a threat to human safety in terms of collisions, and species that may not be included in carcass or crash databases.

Habitat linkage zones can be prioritized by looking at WVC hotspots, using a method referred to as rapid assessment, using expert-based Geographic Information Systems (GIS) models, using GIS movement or population viability models, and/or local knowledge:

Use existing WVC data to identify WVC hotspots. Although WVC data are used as a primary input for many prioritization methods discussed below, using WVC data by itself it is not very forward looking, as changes in land-use patterns can cause changes in animal movements and road mortality. In addition, WVC hotspots ignore locations where animals may be crossing the road successfully, or where animals may want to cross but shy away from the road corridor once they get close (aborted crossing attempt). This is especially important to recognize should the barrier effect of the road be increased through such methods as wildlife fencing. Special attention may be required with regard to road mortality and habitat connectivity for threatened or endangered species.
Rapid assessment of statewide wildlife habitat linkage plans involves:
- gathering pertinent, readily available data including aerial photos, vegetation maps, topography maps, wildlife habitat and range maps, and road mortality information; and
- gathering local experts at a workshop where, with these data available for reference, linkage zones are identified and prioritized through a consensus-based process.
It is important to record information about the linkage zones identified (e. g. , location, species of concern, local agencies and individuals with special knowledge of the linkage zone, major purpose or function of the linkage zone, and a prioritization ranking for the linkage zone). While this approach provides a quick assessment, lack of criteria and weighting rules results in subjective results. In addition, the outcome is heavily dependent on who is invited and not invited to the workshop, the knowledge that the invitees have or do not have, and how strongly their knowledge and opinions are expressed and recognized in a group setting. These issues also apply to other expert-based approaches. For general methodologies for prioritizing habitat linkage zones, contact Bill Ruediger, Wildlife Consulting Resources, wildbill@montana.com
A prioritized linkage-zone map was used in Arizona. A method similar to the rapid assessment technique described above was used to create a map during Arizona's Missing Linkages Workshop. This initial map was refined during six ecoregional workshops. GIS databases were helpful in obtaining some of the data used in prioritizing the linkage zones. For information on efforts in Arizona, contact Steve Thomas, Environmental Program Manager, FHWA Arizona Division, steve. thomas@fhwa. dot.gov.
Expert-based GIS models that utilize available data may indicate good habitat or likely crossing locations. These models do not predict actual movement, but use weighted criteria based on presence or proximity of certain types of features such as vegetation, development and waterways. The weighted criteria are based on expert knowledge, but may not translate into actual locations of where wildlife cross roadways. Examples of this method are listed below:
- The Florida Department of Transportation has created a program to identify and prioritize habitat linkage zones that intersect with highways. ⁶ The purpose is to consider locations for safe crossing opportunities for wildlife on a statewide level in order to restore habitat connectivity and ecological processes. This method uses a rule-based GIS model to assimilate multiple factors such as road-mortality hotspots, riparian areas, greenways, protected conservation lands, and known wildlife movement routes. An output of this effort is shown in figure 3.
- The Idaho Transportation Department identified wildlife connectivity areas using an approach that integrated GIS spatial data, GIS linkage model analysis, and expert workshops to identify areas of interest for mitigation consideration in the southeast corner of the State. ⁷
- A number of other models predict optimal crossing locations for specific species such as Key deer in Big Pine Key, FL, and grizzly bear in the area around Evaro Hill, MT. ⁸ ⁹
Figure 3. High priority ecological zones in Florida. (Source: FDOT)
GIS-based movement models are landscape-scale models used to identify key habitat linkages, evaluate habitat fragmentation resulting from human activities, and identify areas where highways intersect with locations of frequent animal movements. Simulated movement models tend to use resource selection functions that map habitat quality and have rules for simulated movements based on habitat quality and permeability of the landscape concerned. The data used to generate a habitat surface for these models are based on some form of animal "presence/absence" or relative suitability information, usually obtained through radio-collared individuals, but can also be derived from other animal detection methods including DNA sampling, sooted track plates, acoustic surveys, and scat-detection dogs.
Local knowledge has historically been important for wildlife biologists conducting research or managing habitats for wildlife. People who live or work in an area for decades can provide valuable information about where, when, and how wildlife move across the landscape. Local participation in project planning is not only good public relations but provides stakeholders with a sense of project ownership, and builds support for the proposed mitigation measures.

2.4 STEP 3: THINK BEYOND THE ROADWAY

Regional planning for WVCs involves a coordinated effort to safely move wildlife across not only the roadway of concern, but also adjacent roadways/railways. Planning should also factor in likely trends in land use for areas adjacent to the roadway in order to maintain future wildlife movement at safe crossing locations. The challenge is to move from a road-oriented approach (i. e. , a linear element, ignoring the surrounding landscape) to a landscape-based approach (i. e. , a linear element, in the context of its surroundings).

Image of roadway mitigation that may be too close to railroad tracks
Figure 4. Mitigation on the roadway may have only moved the problem to the railroad (copyright: Marcel Huijser).

This type of effort is inherently multi-agency in nature and should involve local land grants, cities, counties, special interest groups, landowners, resource agencies, rail lines, State departments of transportation, etc. An example of this type of thinking is described below.

Along Interstate 90 on Bozeman Pass, east of Bozeman, MT, is a habitat corridor where two areas of the Gallatin National Forest are separated by a transportation corridor (interstate highway and rail line), combined with private land under heavy development. Increased interstate traffic and increased land development each threaten to block animal movements, but any effort to preserve the habitat corridor by a single agency would be ineffective. By prioritizing specific crossing locations, land trusts can focus their efforts on preserving the areas that are immediately adjacent to existing or potential future safe crossing locations, and the Montana Department of Transportation can optimize its investment in mitigation measures on the interstate. For information on the Bozeman Pass collaboration in Montana, contact Deborah Wambach, Montana Department of Transportation, (406) 444-0461, dwambach@mt.gov, or the Bozeman-based non-governmental organization American Wildlands, http://www. wildlands. org.

2.5 SUMMARY

Further information about the planning process can be found at the American Planning Association web site (http://www. planning. org) and from the FHWA (http://www.environment.fhwa.dot.gov/integ/index.asp). The FHWA webpage includes resources such as the Eco-Logical report, providing a process for streamlining the mitigation of wildlife impacts, and numerous training opportunities including a workshop for linking planning and conservation to help communities better understand how to improve this linkage.

To most effectively address all the aspects of WVCs during planning:

Efforts ideally must go beyond prioritizing linkage zones to incorporating them into transportation and conservation planning elements that are developed at both local and State levels of government.
At a minimum, the statewide transportation improvement program should be reviewed annually for reconstruction projects that cross WVC hotspots or habitat linkage zones.
Incorporating WVC mitigations into a reconstruction project at the initial project planning stage will be much less expensive than installing them as separate projects.
Innovative funding sources could also be incorporated into a regional WVC reduction strategy (see chapter 7).

TOC