Why are scanning patterns so variable? An overlooked question in the study of anti-predator vigilance

Anti-predator vigilance describes how animals scan their environments for predators. For mathematical simplicity, theories of anti-predator vigilance have generally assumed that animals initiate scans with a constant probability for each unit of time spent feeding. Initiating scans in such a fashion would produce a wide distribution of intervals between scans. Because so much variation was assumed by theory, observations of variation have received little consideration. We show that the original theories of vigilance contain a conceptual discrepancy between the assumed predator and anti-predator tactics. Modified versions of the theory allow variation only under restrictive circumstances that seem unlikely to apply in nature. Plausible reasons for the observed variation in interscan interval are that theories of vigilance assume the wrong predator tactics or the wrong details of predator detection. These possibilities lead to testable predictions about how distributions of interscan intervals should change under different threat and detection conditions.