tank recorded. Then the illumination was switched 

 to the left end of the tank and the time spent in 

 that half by the fish was recorded. The illuminated 

 half of the tank was reversed end for end every 

 5 minutes over 30-minute periods. The score by 

 which the attraction to light was measured was the 

 cumulative amount of time spent by the fish in 

 the illuminated half of the tank. 



The second routine procedure was used when- 

 ever a group of fish were tested (experiments 5 to 

 16). Ten randomly selected herring were placed 

 in the tank, the attracting light on the right was 

 turned on, and the number of fish present in the 

 illuminated half was counted each minute for 

 5 minutes. The illumination was then switched to 

 the left side, and the fish present in the illuminated 

 half of the tank were again counted each minute 

 for 5 minutes. The lights were, thus, alternated 

 from end to end at 5-minut6 intervals until 30 

 counts had been made. The sum of the 30 counts 

 was the score for groups of fish. A variant of this 

 procedure was used in two experiments (13 and 

 16). Each trial lasted only 1 minute, and counts 

 were made at 15, 30, 45, and 60 seconds. The light 

 position remained the same throughout the period 

 of one trial, but equal numbers of trials were made 

 with the left side illuminated and with the right 

 side illuminated. 



In no experiment were the same fish used in 

 successive trials; after being used once, every 

 fish was returned to a separate holding tank and 

 was not used again for at least 2 weeks. 



Comparisons of scores between only two con- 

 ditions were analyzed statistically with a "t"-test ; 

 comparisons among several kinds and levels of 

 treatment were analyzed with a fixed-model 

 analysis of variance. 



GENERAL BEHAVIOR 



After the fish had become accustomed to the ex- 

 perimental tank their behavior stabilized into one 

 of three general patterns : (1) swimming regularly 

 back and forth fr<Mn one end of the tank to the 

 other (in a loose school, if in a group), (2) re- 

 maining at one end of the tank, or (3) milling 

 about randomly throughout the tank. After the 

 room was darkened and the attracting light was 

 turned on, whatever behavior pattern had been 

 adopted by that particular fish or group of fish 



continued; this behavior tended to obscure the re- 

 sponse of the herring to the attracting light. Al- 

 though a brief fluri-y of activity near the lamp 

 frequently occurred at the begining of each trial, 

 the degree of attraction was not obvious by casual 

 observation ; it oould be demonstrated only by re- 

 fjeated counts over an extended period. The fish 

 siiowed no tendency to gather in any locality of 

 optimum light intensity. 



The exact behavioral mechanisms involved in 

 tlie attraction of fish to artificial lights are not 

 known, although several theories have been pro- 

 posed, ranging from a straightforward positive 

 phototaxis to a conditioned response where light 

 is associated with food. One of the more interesting 

 theories is that of Verheijen (1959 and in press). 

 He suggested that an artificial light creates an un- 

 natural light field which leads to a disoriented be- 

 havior of the fish, because "such simplified visual 

 environments" may not "deliver adequate in- 

 formation to all integration levels involved in the 

 jjerformance of the fish's natural behavior." The 

 net effect is one of trapping fish rather than merely 

 attracting them. 



Close observation of herring in some of my own 

 experiments yielded clues about their behavior in 

 an unnatural light field which may amplify some- 

 what the ideas of Verheijen. The typical behavior 

 of a single herring consisted of the following 

 events : 



(1) Starting from the dark half of the tank, 

 the fish swims slowly at first, then with 

 increasing speed directly toward the light. 



(2) Upon reaching the light, sometimes strik- 

 ing it squarely with its snout, the fish 

 turns more or less broadside to it and pro- 

 ceeds to circle it, or swim to and fro in 

 short courses which lie generally along 

 the circumference of a circle around it. 



(3) This behavior is frequently interrupted 

 by movements which give the impression 

 of confusion or disorientation. Close ob- 

 servation indicates that these movements 

 are, in fact, attempts by the fish to orient 

 its dorsal surface toward the light source. 

 When the light is under water, these at- 

 tempts lead the fish to assume momen- 

 tarily vertical postures or horizontal pos- 

 tures on its side with its back toward the 

 light. 



76 



U.S. FISH AND WILDLIFE SBRVIOB 



