demonstrated that the first minute scores 

 for light-adapted herring were, in fact, sig- 

 nificantly higher even at high temperature than 

 those for dark-adapted herring. Four trials at 15 

 to 17.5° C. were made each day for 4 days; each 

 trial was preceded by a 1-hour period of light or 

 darkness. Counts of the fish were made at 15, 30, 

 45, and 60 seconds after the attracting light was 

 turned on. The end of the tank illuminated, left 

 or right, was alternated with each trial. The mean 

 score for the fish adapted to light for 1 hour was 

 significantly higher (F=32.7, P<0.01) than the 

 mean score for the dark-adapted fish (table 7). A 

 significant bias for one side of the tank also was 

 apparent in this experiment. Such a bias some- 

 times occurred for unknown reasons and made left- 

 right alternation of the illuminated side of the 

 tank a necessary part of the procedure in all 

 experiments. 



This experiment showed that previous adapta- 

 tion to light increases the initial attraction of the 

 herring to light regardless of temperature; on the 

 other hand, the holding effect of the light was 

 weakened at high temperature regardless of the 

 prior adaptation. 



Table 7. — Comparison of scores^ for light attraction of 

 herring during first minute of exposure, in relation to 

 prior light experience (experiment 16) 



Herring 



Location of light 



Light Dark 

 adapted adapted 



Left 31 36 



37 33 



29 25 

 34 21 



Right 23 5 



30 6 

 30 10 

 33 7 



Meanscore. 31 ig 



' Score = sum of numbers of herring counted in the illuminated hall of the 

 tank at four 15-second intervals. Ten herring were used in each test; maii- 

 mum possible score = 40. 



ANALYSIS OF BEHAVIOR IN RESPONSE 

 TO LIGHT 



The attraction of herring to artificial lights is a 

 composite behavior pattern made uji of two gen- 

 eral categories of responses: those that draw the 

 fish toward the light and those that hold the fish 

 under the light's influence. The initial attraction 



82 



seems to be a (usually) positive telotaxis, defined 

 by Frankel and Gunn (1961) as direct attainment 

 of orientation, without deviations, to a source of 

 stimulus as if it were a goal. This response is 

 stronger in some individuals than in otliei-s and in 

 a few may even be negative (away from the light) . 



The holding power of the light, on the other 

 hand, is determined by several, often dissimilar, re- 

 sponses of the fish. One of these is the dorsal light 

 reaction discussed earlier; it holds the fish near 

 the light by interfering with nonnal swimming 

 movements which would lead to escape. Another 

 is photokinesis, where- general activity and swim- 

 ming speed increase with increasing light inten- 

 sity ; this response works in opposition to the dor- 

 sal light reaction, tending to cause dispersal. 

 Adaptation and fatigue probably accompany con- 

 tinued exjDosure to the light and may weaken both 

 of the other reactions. Finally, there is the startling 

 or shock effect of sudden changes in light inten- 

 sity, which may repel the fish, as if by fright. 



The response of fish to light is determined by 

 the way in which conditions influence these be- 

 havioral components. Some of these reactions can 

 be summarized as follows: Temperature affects 

 primarily the degree to which the fish are held un- 

 der the influence of the light, probably through 

 its effect on their activity ; higher temperature in- 

 creases general activity, which in tuni tends to 

 cause dispersal. The position of the light above or 

 below the surface also affects the holding power of 

 the light. Herring are accustomed naturally to 

 light rays directed downward from the surface, 

 and light from a source below the surface is likely 

 to produce orientation which interferes with 

 normal swimming and escape from the lighted 

 zone. Previous adaptation affects primarily the 

 initial attraction to the light, which is stronger in 

 light-adapted fish than in dark-adapted fish. 



An attempt to measure the startle effect of light 

 indicated that whenever the attraction or holding 

 power of the light was strong, the startle effect was 

 less pronounced than when the attracting or hold- 

 ing power was weak. The startle effect was meas- 

 ured by the ratio of tlie number of fish in either 

 side of the tank before the light was turned on to 

 the number present immediately afterwards. A 

 correlation of —0.957 was found between these 

 ratios and the scores for 10 experiments selected 



U.S. FISH AMD WILDLIFE SERVICE 



