Cross or bounce? GPS and video data reveal high road encounter rate and altered movement in a common ungulate

Roads constrain wildlife movement and pose collision risk, yet quantifying fine-scale behavioral responses to roads remains difficult. For ungulates with large home ranges, characterizing how individuals behave during road encounters is essential for identifying where roads function as barriers and for mitigating vehicle collisions. We applied Barrier Behavior Analysis (BaBA) to GPS collar data from 125 white-tailed deer ( Odocoileus virginianus ) across three landscapes in southern Illinois, USA (2023–2025), to classify road encounters into behavioral types. We used mixed-effects logistic regression to examine how encounter type varied with land cover, road category, season, and diel period. We also analyzed over 250 h of camera collar videos to characterize behavior associated with each encounter type. Deer had, on average, 0.05 to 7.61 road encounters per individual per day, with nearly half of encounters involving altered behavior. Altered behavior was more likely during the day, during fawning season, on large roads, and with more forest and agricultural cover. Among the three sites, road impermeability was highest at the most forested site and lowest at the most road-dense site. Camera collar video recordings revealed behavior differences near roads. Specifically, deer were less likely to rest in videos associated with quick cross encounters (rapid road crossings without altered movement) compared to videos recorded away from roads. Roads acted as partial barriers even for a common and behaviorally flexible ungulate, altering movement in nearly half of encounters. Because road barrier strength varied with local land cover, road size, and time of day, impermeability can be resolved to individual road segments and periods when deer are least likely to cross. Camera collar video confirmed that these movement-based classifications reflect genuine differences in behavior, supporting the use of BaBA to target collision mitigation where it will be most effective.