HURLEY: SCHOOL STRUCTURE OF LOLIGO OPALESCENS 



Where measurements of squid length are given, 

 they are of dorsal mantle length from tip of the pen 

 to the tip of the tail. The total length of the squid 

 (including arms but excluding tentacles) is about 

 1.3-1.5 times the dorsal mantle length (Fields 

 1965). 



RESULTS 



Response to Disturbance 



One set of factors that caused changes in school- 

 ing can be grouped under "external disturbances." 

 These included introducing objects (such as a net) 

 into the water near the squid or tapping on the side 

 of the tank. The typical response was for the squid 

 to group more tightly and. in cases where it was 

 not already marked, to increase the degree of 

 parallel orientation. The amount of change in 

 schooling behavior and the temporal characteris- 

 tics of this change depended upon the nature and 

 intensity of the disturbance and upon its duration. 



One attempt at quantifying the stimulus in- 

 volved placing an aquarium air stone in the tank. 

 Pressurized air delivered to this air stone in differ- 

 ing amounts and duration produced a stream of 

 bubbles which could be used as a disturbance 

 stimulus of varied intensity and duration. A small 

 stream of bubbles produced little squid reaction, 

 while vigorous water action due to the bubble 

 stream produced marked changes in behavior. 

 Figure 1 shows the changes in three of the school- 

 ing indices in response to a moderate disturbance 

 caused by turning on the air bubble stimulus. The 

 degree of parallel orientation, which was already 

 pronounced, did not change appreciably. But the 

 squid did draw noticeably closer together. 



Schooling Structure as 

 a Function of Squid Size 



Six squid of nearly equal size were haphazardly 

 taken from the holding tank and placed in the 

 experimental tank. The squid swam in this tank 

 for an hour before measurements were made. With 

 the exception of one experiment, a picture was 

 taken of the squid every minute for approximately 

 1 h. During this other experiment, a picture was 

 taken every 10 s for 10 min. This set of experi- 

 ments was conducted during the daylight hours of 

 two different days. All of the squid used in this set 

 of experiments had been captured on the same 

 night. 



There was a decrease in the mean angular de- 

 viation as the size of the squid increased (Table 1). 

 Since small values of the mean angular deviation 

 index are associated with increased parallel orien- 

 tation, the degree of parallel orientation is 

 greatest in schools composed of large individuals. 

 Even in the case of the small individuals, however, 

 the value of the index does not approach what 

 would be expected if the squid were each orienting 

 in a random direction. In a simulation of 1 million 

 values for six randomly oriented fish. Hunter 

 (1966) found that the mode of the frequency dis- 

 tribution was 69°. 



Although the average values for mean angular 

 deviation do give a measure of average departure 

 from parallel orientation for a whole experiment, 

 they do not give an indication of how variable a 

 particular group of squid is in its orientation over 

 time. For example, an experiment of 30 pictures 

 and an average value of the mean angular devia- 

 tion index of 20° could have had all of the 30 values 

 close to 20°. This would indicate a consistent mod- 

 erate degree of parallel orientation over time. On 

 the other hand, such an average value could also 

 come from a situation where the squid had strong 

 parallel orientation part of the time and were 

 much more loosely oriented the rest of the time 

 (e.g., the index value could have been 10° on 15 

 frames and 30° on 15). This kind of difference can 

 be detected if a measure of the variability of the 

 mean angular deviation index for each experi- 

 ment is calculated. The variability (standard de- 

 viation, SD, Table 1) increased with decreasing 

 squid size, indicating that not only do the smaller 

 squid not orient on the average in as parallel a 

 manner as larger squid, but they are also more 

 temporally variable in their orientation. This dif- 

 ference can also be seen if individual experiments 

 are examined. Figure 2 shows the values for mean 

 separation distance and mean angular deviation 



Table l . — Relationship between average size of Loligo opales- 

 cens and parallel orientation and separation of individuals in the 

 six-squid experiments. Each index was calculated for each 

 frame. 



435 



