What is the best way to estimate vigilance? A comparison of two methods for Gunnison's prairie dogs, Cynomys gunnisoni

Vigilance is important for survivorship and reproductive success and is common and conspicuous within hundreds of species across a diverse array of taxa. Vigilance can involve either scanning for predators (antipredator vigilance) or watching conspecifics (social vigilance). From observations of marked Gunnison's prairie dogs, Cynomys gunnisoni (Sciuridae), living under natural conditions at Valles Caldera National Preserve, New Mexico, U.S.A., we compared results from the two most common sampling methods used by behavioural ecologists to measure vigilance: focal sampling (i.e. observing an individual for a specific amount of time and recording the amount of time spent vigilant) and scan sampling (i.e. observing all individuals in an area at a single moment and recording whether each individual at that moment is vigilant or nonvigilant). Information from different individuals and from the same individuals over time both revealed that estimates of vigilance from scan sampling were consistently and significantly higher than estimates from focal sampling. These differences probably resulted because nonvigilant behaviours were more difficult to detect in scan samples than in focal samples. Our results have important implications for behavioural ecologists who want to make intraspecific or interspecific comparisons of vigilance.