| Sumario: | Many models using vigilance to predict the probability of detecting an approaching predator assumes that prey scanning events should be produced at random. Consequently, the length of intervals among successive scans must follow a negative exponential distribution. We analyzed the scanning behavior of the greater rhea, Rhea americana, which is a gregarious, flightless bird, in eastern Argentina. We investigated whether individual and/or collective scanning departed from random and whether this departure varied with group size. We used two simulation models based on observed scanning sequences to assess the effectiveness of vigilance on the individual and collective level when faced with an opportunistic or stalking predator. The analysis of 59 behavioral sequences of wild greater rheas foraging solitary or in groups of two to six or more individuals revealed that the inter-scan length of individual sequences significantly departed from random. In contrast, inter-scan intervals for collective vigilance were shorter than individual ones, but only fit the random expectation for groups of two and five individuals. Models showed that collective vigilance could increase the probability of detecting a predator, thereby reducing their vulnerability, independent of whether the predator uses a stalking or opportunistic approaching strategy. © Japan Ethological Society and Springer 2008.
|
|---|