Background and Justification
The reconstruction of the MRL overpass structure on I-90 at approximate milepost 314.1, provides an excellent opportunity to install wildlife fencing. The installation of this fencing may reduce anmial vehicle collisions, reduce carcass removal requirements on maintenance, and improve the habitat linkage of this corridor. A fence addition to the MRL bridge rebuild is designed to serve as a pilot study to prevent wildlife from crossing over Interstate 90 in an area with the highest rates of animal-vehicle collisions found between Bozeman and Livingston, and re-direct wildlife in transit so that they can safely pass underneath the highway using 2 existing bridges and 3 existing culverts,

Design Philosophy
These recommendations are based on using wildlife proof fencing and related appertanences to cause animals to cross I-90 either under the Interstate using existing structures or at segments of the Interstate with good sight distance (i.e., tangent segment with relatively level side slopes) where drivers have greater chance of avoiding collision. Field visits were made with MDT engineers and with State deer biologist Dave Pac to determine the best placement of structures. Dave Pac recommended a shorter length of fence to prevent deer from crossing the highway only in areas where sight distance is poor. He feels that deer may not travel more than about ½ mile along the fence to search for a way across. However, a longer fence design may be more effective at redirecting other species such as bears, elk, and moose. Two basic alternatives are presented in this report : a deer optimum with shorter fencing, and a longer version which would keep more animals off the highway, but may reduce deer use of habitat and hinder migration.

Design Recommendations
This design recommends wildlife fencing, ungulate guards, landscaping, and end treatments (angled fence, cobble, and jumpouts). The general design for the fencing and ungulate guard elements are discussed below.

Fencing
Two general types of wildlife exclusion fences are commonly used: 8.8cm (4-inch) wire mesh (Clevinger et. al.) as used in Banff National Park, Alberta, Canada and in the Vancouver Island Highway Project in British Columbia, and cyclone mesh (Alaska DOT) as used along the Glenn Highway in Alaska. Either type has proven effective, but the 4-inch mesh (Figure 5) will have less visual impact to motorists on the highway. Fence should be installed with 4-inch diameter posts spaced about 4.6 m (15 ft) apart. Wood posts should be 4.0 meters in length direct buried 1.6m. Corner posts may be placed in concrete for additional strength. The fence will generally be installed in place of or directly adjacent to the existing right-of-way fence. The wire mesh should go all the way to the ground: any gaps larger than six inches in height should be filled in with soil. At the location of the existing double 2.5m-high culvert at approximate milepost 314.4 the fence should continue up and over the top of the culverts to allow animal use of the culverts to cross the highway.

Ungulate Guard
Normal cattle guards are ineffective for moose, bison, elk and other large ungulates. Guards with larger openings have been effectively used in Canada where they have been very successful. The British Columbia Ministry of Transportation has found that 8-foot high wildlife exclusion fencing installations constructed with ungulate guards reduce wildlife accident rates by 97% to 98%. The two main examples are the Coquihalla Highway and the Okanagan Connector. This setup has been installed with dedicated wildlife underpasses. (Sielecki, pers. Comm.). We feel that the existing underpasses along I-90 will also act as effective transit routes if animals are re-directed by fencing and ungulate guards. We have requested drawings of these structures from the British Columbia Ministry of Transportation (BCMOT) and will provide them to MDT.

Landscaping
It may be possible to provide cover for animals near the entrances to culverts and bridges in some instances so that they feel more secure in attempting to cross the highway through these structures. Palatable plant species could be incorporated near bridges and culverts to entice ungulates to browse in their current location instead of looking for food on the other side of the freeway. However, Dave Pac, Montana Department of Fish Wildlife and Parks, recommended that areas under bridges/overpasses be left as open as possible in order to encourage deer to pass through. Currently the culverts are difficult for animals to access because of thick brush and rip-rap. It would be possible to clear brush and landscape the approaches so that animals can find and access the culverts more easily. In general, we would recommend removing brush and shrubs for deer; this would probably benefit other species as well.

End Treatments
One challenge with wildlife fencing is animals making end-runs around the fence and becoming trapped between the two fences on the right-of-way. Where possible the fence should abut directly against cattle/moose guards, the corners of bridge abutments, and existing terrain elements that limit animal movement. In other cases the use of cobble, flared fence lines, and jump-outs will be used to limit end-runs around the fence.

Cobbles have been found to deter animal movements; they prefer to travel on more solid surfaces if possible (Leonard and Bissonnette 1997). Boulders and cobbles should be small enough that they do not pose a hazard to motorists that might run into them, but large enough to cause an impediment to wildlife. 20-30cm diameter rounded river cobbles were used in a fencing project near Park City, Utah (Leonard and Bissonnette 1997). Cobble beds should extend a short distance (50 feet).

The specific angle and length of the fence flare will depend on the surrounding terrain, but will typically be a 30-45 degree angle for 10-12 feet.

If animals still make an end-run around the fence it is a good idea to provide a one-way exit near the end treatment. There are two general types of fence modifications that allow animals trapped inside the fence on the roadway to escape. Jump outs are ramps made of earth contained in berms that allow an animal to walk up to the top of the ramp where the fence is open and then jump about 6 feet to the ground outside the fence. One-way gates have mechanical barriers that allow an animal to pass in one direction only. Gates tend to get bent and damaged by wildlife and typically require more maintenance than jump-outs. Jump outs are to be placed approximately 100 meters from locations where animals can make end-runs around the fence. We have requested drawings of wildlife gates and jump-outs from the BCMOT and will provide them to MDT.

Figure 7 shows the two typical end treatments to be used. The second end treatment is used where space and terrain do not allow for the first option. An alternate arrangement is shown in Figure 8.

Figure 7: Typical End Treatments 1 & 2

Figure 8: Alternate End Treatments 1 & 2

MDT assistance has been invaluable for consultation on design modifications and options. Research staff discussed site-specific mitigation options with MDT biologists, engineers and administrators to explore the feasibility of incorporation into highway design for increased wildlife use. Field visits to the study area with MDT personnel were conducted during the winter of 2002.

MDT will need to be an integral part of the proposed wildlife exclusion fencing project, but MDT's involvement will be limited primarily to construction and maintenance of the wildlife fence and installation of the ungulate/cattle guards. Fencing recommendations entitled: "Design Recommendations for Wildlife Fencing on I-90 Bozeman Pass To be incorporated with: Project IM 90-6(88)314 MRL Overpass 7km E of Bozeman" have been provided to Jason Giard, District Administrator of the Butte District of MDT. For the preferred alternatives, this will require the installation of about 1.2 miles of fence (Options A-1 plusB-1 plus C-2 in the Design Recommendations supplied: see Appendix). Although the fence should greatly reduce animal-vehicle collisions in the project area, some maintenance work may still be required to remove animal carcasses. Once the fence is installed, CERI and WTI staff will monitor wildlife use of the area to determine the effects of the fencing in relation to monitoring data collected before the fence was installed.