Development and evaluation of devices designed to minimize deer-vehicle collisions : phase II.

Osborn, David A.; Gulsby, William D.; Stull, Daniel W.; Cohen, Bradley S.; Warren, Robert J.; Miller, Karl V.; Gallagher, George R. · 2010 · ROSA P / Daniel B. Warnell School of Forestry and Natural Resources.

archive: archived pipeline: cataloged verified

Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)

Summary

This report details Phase II of a collaborative research project funded by the Georgia Department of Transportation to evaluate devices designed to minimize deer-vehicle collisions (DVCs). Motivated by the failure of sight- and sound-based deterrents in Phase I, the study focused on testing the efficacy of physical barriers, specifically various fencing designs and tactile barriers, to restrict deer access to roadways. Additionally, the project aimed to characterize the visual capabilities of white-tailed deer to inform future mitigation strategies. The researchers conducted experiments using both captive and free-ranging white-tailed deer. In captive trials, they evaluated the effectiveness of woven-wire fences ranging from 1.2 to 2.4 meters in height, opaque fencing, a layer of Type III rip-rap rock, and a prototype outrigger fence consisting of a 1.2-meter wire fence with 0.6-meter outriggers angled at 45 degrees. To assess real-world application, the team constructed 3.2 kilometers of fencing along roadways at Berry College, comprising 1.6 kilometers of 2.4-meter woven-wire fence and 1.6 kilometers of the outrigger design. They monitored 14 adult does fitted with GPS collars to compare home ranges, core areas, and crossing behaviors before and after fence installation. Furthermore, the team developed an automated "deer-training-device" to study deer vision by training captive deer to associate specific light stimuli with food rewards. The findings indicated that woven-wire fences shorter than 1.8 meters and opaque fencing of similar heights were ineffective at excluding deer. Rip-rap rock also failed as a tactile barrier, as deer crossed it easily and vegetation quickly established among the rocks. In contrast, woven-wire fences taller than 2.1 meters and the 1.2-meter fence with a top-mounted outrigger angled toward the deer were highly effective. In the free-ranging study, the 2.4-meter fence reduced deer crossings by 98%, while the outrigger design reduced crossings by 90%. Although fencing did not alter deer home range sizes, deer frequently circumvented fence ends. The outrigger design was highlighted as a promising mitigation tool due to its lower cost and potential to function as a one-way barrier, allowing deer to exit roadways without becoming trapped. Regarding vision, the automated training system successfully conditioned deer to discriminate between light stimuli with greater than 75% accuracy, validating the method for further visual threshold research. The study concludes that while tall woven-wire fences are effective, they are expensive and risk trapping deer if they circumvent ends. The outrigger fence offers a viable, more affordable alternative that may reduce DVCs by permitting one-way travel away from the road. The research underscores the importance of using localized data on deer home ranges to determine the necessary length of fencing to prevent circumvention. Additionally, the successful development of the automated training device provides a foundation for future studies on deer visual perception, which could lead to innovative, vision-based deterrents for roadway safety.

Key finding

Fence crossings by free-ranging deer were reduced by 98% for 2.4-m woven-wire fences and 90% for outrigger designs, while shorter fences and rip-rap proved ineffective.

Methodology

mixed_methods

Sample size: 14

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• rail grade crossings