CAHN: SENSORY FACTORS 



doubt involved in lateral vision, with a perhaps 

 different cone distribution. Nakamura did not 

 study this for kawakawa, but others have dem- 

 onstrated that schooling fish are nearsighted in 

 forward vision but farsighted in lateral vision 

 (Baylor and Shaw, 1962). This suggests that 

 perhaps the kawakawa in our experiments with 

 partitions shifted spacing and position to make 

 more use of lateral vision. A possible reason 

 may be that in the absence of mate-generated 

 hydrodynamic cues, nearsighted vision is inad- 

 equate for efficient tracking of their schoolmates. 

 To explain all of these visual factors, more in- 

 formation is needed on such problems as reso- 

 lution and cone density in the different retinal 

 regions of kawakawa. 



One other question relative to vision requires 

 some comment: did the partitions produce re- 

 flections sensed by the fish which led to the 

 changes observed? As previously noted, fish in 

 a school in the presence of a partition positioned 

 themselves in different locations relative to the 

 barrier, depending on the fish-to-fish distance 

 across the partition, rather than on the distance 

 from the partition (Figure 4). Thus if there 

 were any reflections, they did not appear to be 

 of any significance. The schooling tendency pre- 

 dominated: relative to this it should be pointed 

 out that it is unlikely that the change in spacing 

 could be attributed, to any major extent, to a 

 waning of the stimulus to school. Fish-to-fish 

 orientation sometimes persisted across the par- 

 tition for as long as nine days. 



We have considered the prime sensory factors, 

 and omitted tactile and chemical contributions. 

 Since these fish rarely touch each other when 

 swimming except under special circumstances 

 of school structure and size (Breder, 1967), 

 there is no need for further concern with the 

 tactile sense. Chemical cues, in our opinion, ' 

 were probably not transmitted rapidly enough 

 across or through the partitions to be of any 

 consequence. Even with holes in the partitions, 

 as in one series of tests, water-borne transfer 

 was very slow. Also, the part played by chemo- 

 reception in fish schooling (McFarland and Moss, 

 1967; Moss and McFarland, 1970) appears un- 

 important for obligate schoolers such as tunas 



and other scombroid forms (the use of obligate 

 here is as in Breder, 1967). 



The greater variability noted in the side-to- 

 side spacing in the presence of a partition re- 

 flects, in our opinion, what happens when the 

 normal hierarchy of sensory systems is inter- 

 fered with. The dynamic stability of a fish school 

 depends on feedback from all of the diflferent 

 senses. The partitions disrupted this balance, 

 and although schooling persisted, certain lim- 

 itations were imposed on the process. 



In conclusion, this study suggests that the di- 

 agonal fish-to-fish positional orientation and the 

 side-to-side spacing assumed during schooling 

 may be more essential for hydrodynamic detec- 

 tion than for visual cues. 



ACKNOWLEDGMENTS 



Thanks are extended to the Honolulu Biologi- 

 cal Laboratory of the National Marine Fisheries 

 Service, and also to the University of Hawaii 

 and the National Science Foundation, for partial 

 funding of this work. Technical assistance of 

 the staflF at the Honolulu Biological Laboratory 

 is gratefully acknowledged, especially that of a 

 former staflf member, Dick Holloway. Dr. Wil- 

 liam Siler, Downstate Medical Center, State Uni- 

 versity of New York, Brooklyn, N.Y., made 

 available facilities for the computer analysis of 

 digitized film data, gave statistical advice, and 

 read and criticized the manuscript. Dr. John 

 Hunter, National Marine Fisheries Service Fish- 

 ery-Oceanography Center, La Jolla, Calif., and 

 Dr. John Magnuson, Laboratory of Limnology, 

 Department of Zoology, University of Wisconsin, 

 Madison, Wise, read the manuscript and oflFered 

 valuable suggestions. 



LITERATURE CITED 



Baylor, E. R., and E. Shaw. 



1962. Refractive error and vision in fishes. Science 

 136: 157-158. 

 Breder, C. M., Jr. 



1959. Studies on social groupings in fishes. Bull. 

 Amer. Mus. Nat. Hist. 117: 393-482. 



1965. Vortices and fish schools. Zoologica 50: 97- 

 114. 



1967. On the survival value of fish schools. Zoo- 

 logica 52: 25-40. 



203 



