ABSTRACT Baiting is widely used to improve the capture and observation of invasive mesocarnivores, but it can also alter behaviour and distribution, potentially increasing ecological risks, such as non‐target interactions and pathogen transmission. However, responses to baiting across temporal scales remain poorly understood. We investigated scale‐dependent responses of invasive raccoons ( Procyon lotor ) to baiting in a post‐nuclear‐disaster agricultural landscape in Fukushima, Japan. We used cafeteria‐style bait trials, in which multiple bait types were offered simultaneously at each station; together with GPS telemetry, we evaluated responses from bait choice to short‐, mid‐ (daily), and long‐term space use. Individual bait preferences, short‐term movement responses, daily responses and long‐term space use were analysed using Plackett–Luce models, hidden Markov models, minimum distance to trap sites and autocorrelated kernel density estimation, respectively. Seven GPS‐collared raccoons visited bait stations, yielding 15,837 GPS locations. Bait choice varied markedly among individuals. Across trials, dry cat food (0.40) and caramel corn (0.38) had the highest first‐selection probabilities, whereas dried ramen noodles were seldom selected first (0.01). Some individuals consistently preferred other bait types such as sake lees. Short‐term analyses showed strong attraction to bait stations, with area‐restricted search‐like movements increasing from 22.2% during non‐baiting periods to 53.8% during baiting. At the daily scale, raccoons moved closer to trap sites on baiting days, indicating increased proximity to management locations. In contrast, all bait stations fell within long‐term home ranges and often within core areas, suggesting localised activity intensification rather than broader restructuring of space use. Overall, baiting induced strong but scale‐dependent behavioural responses, with localised increases in site use but little evidence of broader restructuring of long‐term space use. These findings suggest that the effectiveness and interpretation of bait‐based monitoring and management may depend on both temporal and individual‐specific bait preferences.
彰 et al. (Mon,) studied this question.