Is social coordination during escape flights a general phenomenon in birds?

Abstract

Escape from predatory attack as a socially coordinated group is observed in many social animals, including birds, especially those in more open habitats where the group itself may be the only source of protection from an attacking predator. For many social birds, however, woody vegetative cover is the main refuge from attack, but such birds might nevertheless benefit from social coordination during escape flights to cover. Such benefits could reflect the confusion effect, selfish herd effect, or the simple dilution of risk. We examined this possibility of coordinated escape in mixed flocks of wintering passerellid sparrows (Passerellidae). These free‐living birds fed in a patch of food flanked on opposite sides by two refuges composed of woody cover. Under such conditions, coordination in escape behavior should be expressed as a tendency to escape together as a group to the same cover location. Such behavior, however, was not the rule. During spontaneous flushes to cover, a group of escaping birds stayed together only when one cover location was clearly closer than the other. With cover equidistant from the food patch, escaping flocks tended to split about evenly between cover locations. Birds in close proximity prior to an escape flight did not show enhanced escape coordination, nor did those feeding at significant distances from protective cover. Evidence of escape coordination was observed in small groups (two–four birds), but even in such groups, flock splitting during escape was generally the rule. Flock splitting during attacks might reflect some sort of strategic decision‐making process that lessens the risk of capture, but the most parsimonious explanation is that (all else equal) birds head for the nearest refuge, largely irrespective of the behavior of their flockmates. Our results thus provide little evidence of flock‐wide social coordination during escape flights in cover‐dependent birds.