FISHERY BULLETIN: VOL 76, NO. 2 



slight backward and forward motion caused by jets 

 of water from the siphon. Even when disturbed, 

 the squid do not make long extended swims which 

 would tend to sort out those of differing swimming 

 ability. In the field, the most common response of a 

 squid school to a disturbance (the presence of a 

 scuba diver or a shark) is to clump closer together 

 and move off a slight distance. On one occasion 

 when I was diving in a large spawning school 

 (several thousand individuals), the squid executed 

 the same type of maneuver that has been reported 

 for fish schools. Instead of moving off, the school 

 completely enclosed me, leaving a spherical space 

 of approximately 3 m radius around the "pred- 

 ator." 



One other piece of evidence suggests that it is 

 not differences in swimming speed alone which 

 cause the squid to school according to size. While 

 diving in the Bahamas in the Hydrolab underwa- 

 ter habitat, we observed a school of squid which 

 routinely visited the habitat. This school was 

 composed of Doryteuthisplei, a species which quite 

 closely resembles L. opalescens and presumably 

 has similar swimming ability. This school con- 

 sisted of seven squid and, in this case, was not 

 composed of individuals of the same size, the 

 largest individual being at least two times the 

 length of the smallest individual. We chased this 

 school several times but were never able to force 

 them to separate. The smallest squid maintained 

 the same swimming speed as the largest squid. 



It is possible that squid maintain schools of in- 

 dividuals of a fairly narrow size range because of 

 social factors. Generally, workers studying school- 

 ing have assumed that all of the fish in a school 

 may be treated as equivalent individuals in the 

 production of the behavior and that there is no 

 social structure in the schools. In fact, some work- 

 ers have suggested that schooling is really just a 

 modified form of individual cover-seeking be- 

 havior (Williams 1964; Hamilton 1971). This as- 

 sumption of equality of individuals may be an 

 untenable one for squid schools. In the field, 

 Hochberg and Couch ( 1971 ) observed signaling by 

 some members of a school of Sepioteuthis sepioidea 

 which they felt prevented other squid from joining 

 the school. Furthermore, in the laboratory, I have 

 observed complicated agressive interactions in L. 

 opalescens which certainly demonstrate that all 

 squid cannot be considered behaviorally equiva- 

 lent individuals at all times (Hurley 1977). 



One aspect of schooling in fish which has been 

 emphasized by many workers is that the structure 



of schools may change as a function of the age or 

 the physiological condition of the fish. Van 01st 

 and Hunter ( 1970), for instance, found that in five 

 species of marine fishes, schools of young fish were 

 less compact and showed greater differences in 

 angular headings than did schools of adult fish. In 

 addition. Hunter ( 1966) showed that distances be- 

 tween jack mackerel tended to increase with food 

 deprivation, while Keenleyside (1955) noticed 

 that sticklebacks were more densely packed in a 

 school when well fed than when starved. 



I attempted to determine whether similar 

 phenomena were observable in squid schools. 

 Schools of small squid ( 7-9 cm mantle length) gave 

 the impression of being less cohesive than schools 

 of larger squid ( 13-15 cm mantle length). This was 

 supported to some extent by the quantitative mea- 

 surements, particularly those of angular orienta- 

 tion. The variability was also higher for all of the 

 indices for the smaller squid. It has been suggested 

 for fish (Van 01st and Hunter 1970) that the ob- 

 served change with size could have been an adap- 

 tation to the higher food requirements of the 

 juvenile fish. This speculation is supported by ob- 

 servations that a number of species school less 

 cohesively under conditions of food deprivation. 

 The same explanation may also hold for squid, but 

 my existing data do not support it. I ran two exper- 

 iments in which schools of six squid were filmed 

 before and after feeding. In one experiment, there 

 were no significant differences in the schooling 

 indices before and after feeding, while in the other, 

 there was significantly less school cohesion and 

 parallel orientation after feeding. In any event, it 

 is not possible to guess which factors are instru- 

 mental in this increased cohesiveness and consis- 

 tent geometry. 



As is the case for fish, vision seems to be the 

 primary sensory system used in squid schooling. 

 Squid will readily school across a clear, rigid 

 Plexiglas barrier, although they tend to stay 

 somewhat farther apart than they normally 

 would. Investigators dealing with fish also found 

 that the presence of a clear, rigid barrier caused 

 abnormalities in the spacing between individuals, 

 in some cases increasing the fish-to-fish distance 

 ( Cahn 1972) and in some cases decreasing it (Shaw 

 1969). These workers speculated that this change 

 was due to lack of lateral line input and a resultant 

 loss of information concerning the position of the 

 adjacent fish. Squid do not have a similar exten- 

 sive vibration-sensitive system, although they 

 may be able to detect vibrations with their stato- 



440 



