How does a flock of birds wheel and swoop in unison?

This is from the Straight Dope whose motto is:
Fighting Ignorance since 1973 (It's taking longer than we thought.)

March 11, 2004

The highly coordinated movements of flocks of birds or schools of fish are among the most fascinating phenomena to be found in nature. The group seems to turn and maneuver as a single unit, changing direction almost instantaneously, leading some researchers to hypothesize that electromagnetic communication or even "thought transference" must be involved. In reality this behavior results from far less mysterious causes. Such movements are a prime example of emergent behavior: the behavior is not a property of any individual bird, but rather emerges as a property of the group itself. There is no leader, no overall control; instead the flock's movements are determined by the moment-by-moment decisions of individual birds, following simple rules in response to interactions with their neighbors in the flock.

First, I should mention why animals aggregate in flocks, herds, and schools in the first place. While there are many reasons, the most pervasive seems to be that it serves as a defense against predators. Having many eyes together ensures that at least some will spot a predator while others are feeding, snoozing, or looking in the wrong direction. Once the group takes flight, the predator may have trouble focusing on a single target and become confused. It may also be physically dangerous for a predator to plunge into a seething mass of prey. In some cases, larger or more aggressive prey species may be able to offer a coordinated defense and fend off a predator that would make short work of an isolated individual.

There can be other benefits of flocking, such as locating clumped food resources or ensuring accurate navigation on migration or to roosts and local feeding areas. Some species may also aggregate for social and reproductive reasons. In the special case of formation flying by large birds such as geese and pelicans, there is an energetic benefit, since following birds can take advantage of vortexes in the air produced by the ones ahead of them. (Although such formations clearly have leaders, these are temporary ones. Because a lead bird does not gain any energetic advantage from its position, it will drop back after a time while another takes the lead. Flock members probably do not do this on any regular rotation, although it's possible that larger and stronger birds are in the lead a greater percentage of the time.) However, none of these functions seems to come into play as generally as the anti-predator one.

When frightened by a flying predator, a flock of small birds such as sandpipers or starlings will bunch up and fly in as compact a mass as possible. A dive-bombing falcon will avoid plunging into such a crowd for fear of injuring itself in a collision, but instead will seek to pick off laggards or birds shearing away from the flock. The flock itself will veer and turn in erratic fashion, making it difficult for the predator to predict its movements.

Observation shows that there are no leaders (at least not for more than a few seconds at a time), since different birds will be at the front of the flock every time it changes direction. Research by Wayne Potts, published in the journal Nature in 1984, helped explain how flock movements are initiated and coordinated. Potts, through a frame-by-frame analysis of high-speed film of sandpiper flocks, found that any individual can initiate a flock movement, which then propagates through the flock in a wave radiating out from the initiation site. These "maneuver waves" could move in any direction through the flock, including from back to front. However, the flock usually only responded to birds that banked into the flock, rather than away from it. Since birds turning away from the flock run the risk of being separated from it and getting picked off by the predator, others will not follow them. Besides its obvious benefits for individuals, this rule helps prevent indecision by the flock and permits it to respond rapidly to attack.

Once one of these waves began, Potts found that it spread through the flock far more rapidly than could be explained by the reaction times of individual birds. A bird's mean startle reaction time to a light flash as measured in the laboratory was 38 milliseconds, but maneuver waves spread through the flock between birds at a mean speed of less than 15 milliseconds. However, the first birds to respond to an initiator took 67 milliseconds to react. Potts proposed that birds farther away from the initiation site were able to see the wave approaching them, and could "get set" to respond before it actually reached them. He dubbed this the "chorus line hypothesis," in analogy to Rockettes at Radio City Music Hall who can see and anticipate an approaching high leg kick when it is still well down the line. Films of human chorus lines show that rehearsed maneuvers, initiated without warning, propagate down the line at less than 108 milliseconds, almost twice as fast as the human visual reaction time of 194 milliseconds.

When flocks are not under attack, but instead leaving a roost site to go to a feeding area, they may also swerve back and forth apparently aimlessly, because random movements by single individuals can easily generate changes in direction. However, eventually a sort of consensus will develop based on the motivation of the majority of the flock members, and the flock will fly off to its destination in a fairly direct manner.

Because complex behavior can be generated by the application of simple rules, flocking behavior has been a favorite subject for computer modeling. One of the best known efforts of this kind has been the creation of "boids," generic flocking computer organisms, by Craig Reynolds in 1986. Reynolds' three fundamental "laws" of flocking are: (1) separation - steer to avoid crowding local flockmates; (2) alignment - steer towards the average heading of local flockmates; and (3) cohesion - steer to move toward the average position of local flockmates. Computer-generated flocks of bats and penguins based on boids were first used in the film Batman Returns in 1992, and have appeared in many films subsequently.

References

Potts, Wayne K. 1984. "The chorus-line hypothesis of coordination in avian flocks." Nature 24: 344-345.

Boids: http://www.red3d.com/cwr/boids/

(includes many additional references)

— Colibri