Abstract
<jats:p>Urban habitats can serve as wildlife refuges, or as ecological traps. As global urbanization continues to increase, native prey species may experience reduced threat from natural predators that tend to avoid human-populated areas. However, such protection often comes at the expense of susceptibility to synanthropic predators, one of the most prevalent being domestic house cats (Felis catus). While cats are a known driving force of mortalities in urban wildlife, their sublethal effects remain understudied. To better understand the behavioral effects of synanthropic predators on native urban prey, we examined responses of western fence lizards (Sceloporus occidentalis) to Artificial Intelligence (AI)-generated visual stimuli of native predators (snakes) and synanthropic predators (house cats). Visual stimuli included: California striped racers (Masticophis lateralis), a native predator; domestic cats, a synanthropic mammalian predator (Felis catus); and desert cottontails (Sylvilagus audubonii), a native mammalian non-predator. We quantified lizard responses as the change in rates of looking and locomotion during and following the video presentation. Our results revealed that lizards discriminated cats from the control by looking more within the first 90 seconds following exposure. Lizards looked more across the entire 5-minute observation period following exposure to all stimulus types compared to control conditions, indicating that AI generated visual stimuli effectively elicit behavioral responses. Together, these findings indicate that urban lizards initially detect and respond to visual cues of cats, suggesting that cat-induced mortality is not a result of inability to detect such predators. Thus, cats influence urban prey not only through direct predation risk, but also changes in behavior associated with the presence of synanthropic predators.</jats:p>