Abstract Field manipulations of perceived predation risk are frequently used to interpret changes in gaze shift patterns between foraging and anti-predator vigilance. Gaze shifts relate to spatial attention mechanisms studied in psychophysical and neuroimaging laboratory studies. However, connecting laboratory-based insights to naturalistic contexts involving predation risk remains challenging. To bridge this gap, we developed a study of Florida scrub-jay sentinels (Aphelocoma coerulescens). Sentinel bouts exclude foraging, providing a simplified focus for studying attention in the wild. We first defined a neurocognitive agent-based model. For initial model validation, we manipulated background predation risk in simulations, which produced head rotation behaviour consistent with empirical literature. We then conducted an experiment both in the field and in computational simulations based on the model. In the field experiment and its simulation, we manipulated perceived acute predation risk and measured a decrease in head rotation frequency. The model suggests that greater background risk requires more frequent head rotations to enhance predator detection, whereas greater acute risk requires more observations of fine-grained (possible) predator locations per head position. This shift from detection to localization is consistent with a shift from alerting to orienting attention. Our approach demonstrates a promising path for integrating ecological field experiments with laboratory-based comparative (neuro)cognition research.
LINSON et al. (Wed,) studied this question.