Mechanism

Why do schools consist of similar fish?

One might assume upon observing the makeup of many schools that fish self-segregate into groups by species, age, and other shared traits. According to Landa 1998, however, schools have no incentive to discriminate against outsiders; in fact, there may be an incentive to include fish of a different species, age, or appearance, as these outsiders stand out and draw the attention of predators off of the rest of the school. For this same reason, outsiders have a disincentive to attempt to join schools into which they will not blend, as predators have been observed to preferentially select fish which have a different appearance or size from the others in the school (Landa).

It is interesting to note that kin recognition, which has been shown to occur in fish, has no effect on the makeup of fish schools. Griffiths and Magurran 1999 arranged for related guppies to be split such that their reaction to kin with which they had been raised could be compared to their reaction to kin with which they had never before interacted. They showed no preference towards schooling with unfamiliar kin, indicating that the makeup of schools is determined by familiarity rather than by kinship.

kinship schooling
Time schooling with familiar and unfamiliar kin and non-kin. Figure from Griffiths and Magurran 1999.

How are schools maintained in spatial proximity?

It has been proven as early as 1976 that visual recognition is not the only method of spatial organization in schools. Pitcher et. al 1976 determined that fish whose vision had been temporarily removed by means of blinders (made in this case of exposed film) were able to school relatively normally, although their reaction time was diminished.


Image source: Pitcher et. al 1976.

This figure from the Pitcher study shows the effect of visual stimuli (a sudden hand motion) on a school and the failure of a blinded fish to respond to it. The authors note that unless a blinded fish was within one body length of its neighbors, it would not pick up on the change in direction, and that even if it was, it would respond reactively to the movement of the other fish, which created a serious lag in response time. They account for this ability to nonvisually sense neighbor position as a function of the lateral line organ. This accounting was verified by a subsequent experiment in the same paper, in which fish received cuts through the lateral line organ. Those fish which had both cuts in the organ and blinders were unable to school at all, while those which recieved cuts in the organ but retained their vision schooled relatively normally.

It has been firmly established (Montgomery et. al 2003) that the lateral line organ functions by means of sensing water motion, and that fish are able to utilize this data for actions as complex as maintaining station underwater by positioning behind an object which creates an area of low current (object entrainment) and capturing prey without light; it is likely that this is the mechanism by which fish determine the hydrodynamic characteristics of their schooling structure (see Adaptive Value).