quately monitored in the other experiments with 

 oxygen concentration, observations were made on 

 groups of 10 fish 15, 30, 45, and 60 seconds after 

 the attracting light was turned on. The location 

 of the light was alternated between right and left 

 with each trial. Four trials were made each day; 

 normally siiturated and supersaturated water were 

 used on alternate days. The sequence was repeated 

 twice, making eight trials at each oxygen level, all 

 at a mean temperature of 15.5 to 17.0° C, and mean 

 oxygen concentrations of 7.3 p.p.m., 88 percent 

 saturation and 16.8 p.p.m., 200 percent saturation. 

 The mean scores of 31 and 29 were not significantly 

 different. 



The weight of evidence indicates that neither 

 oxygen concentration nor percentage saturation 

 has any effect on the attraction of herring to light. 



EFFECT OF PREVIOUS ADAPTATION 



Experiment 14. Preliminary observations showed 

 a tendency for herring to be le&s strongly attracted 

 to light if they had been kept in darkness before- 

 hand. Therefore, for most of the experiments the 

 herring liad I)een held in full room illumination 

 so that their response would be as strong as pos- 

 sible. Nevertheless, some specific tests seemed de- 

 sirable to confirm the preliminary observations. 



Sixteen trials, two each day, were made with 

 groups of 10 herring at a temperature of 5° C. 

 Each group was exposed to a medium-intensity, 

 underwater light after a 3-hour period of adapta- 

 tion to light or darkness just before each trial. 

 These adaptation periods were alternated with re- 

 spect to time of day, forenoon or afternoon. 



Because the fish had necessarily to be held at all 

 times other than tlie 3-hour adaptation period at 

 some lighting condition or another, the possibility 

 existed that whatever this lighting was would also 

 influence the subsequent behavior of the fish. 

 Therefore, half of the trials were preceded with 

 exposure to total darkness the night before and 

 half with full illumination. The 10 trials were ar- 

 ranged as follows: eight in the morning and eight 

 in the afternoon; four of each of these eight were 

 preceded with a 3-hour period of darkness, and 

 four with a 3-hour period of light ; two of each of 

 these four followed an overnight period of dark- 

 ness and two an overnight period of light. An 

 analysis of vaiiance showed significant variation 

 only witli respect to the 3-hour pretrial light- or 



dark-adaptation period (F = 5.8, P = 0.05). The 

 effects of overnight lighting and time of day were 

 of doubtful significance (F = 3.4, P = 0.1, and 

 F = 2.6, P>0.1, respectively) . The scores are shown 

 in table 6. Pretrial adaptation to darkness reduced 

 the effectiveness of light attraction. 



Table 6. — Comparison of scores ' for light attraction of 

 herring in relation to prior tight experience and time of 

 day (experiment 14) 



Time of day 



Overnight 

 darlraess 



Overnight 

 light 



AM 



PM.. 



Mean score 114 143 



• Score =suni of numbers of herring comited in the illuminated half of the 

 tank at 30 1-minute intervals. Ten herring were used in each test; maximum 

 possible score =300. 



2 Light condition for 3 hours preceding trials. 



Experiment 15. This experiment was done later 

 in the year than experiment 14 at the higher tem- 

 perature range of 15 to 17.5° C. No allowance was 

 made for any previous light experience of the fish 

 prior to a 6-hour light or dark period of adapta- 

 tion before each trial. The experiment also in- 

 cluded the additional variable of oxygen concen- 

 tration and was actually a part of experiment 12. 

 The mean score for 15 trials preceded by a 6-liour 

 dark period was 122; that for 15 trials preceded 

 by a 6-hour light period was 136. The difference 

 between them was not significant (F = 1.48, P> 

 0.2). Apparently, previous adaptation to light or 

 dark makes little difference in the response of her- 

 ring to light at high temperature. 



Experiment 16. The pretrial adaptation of her- 

 ring to light produced the strongest positive re- 

 sjwnse when the trials were made at low tempera- 

 ture. The most marked attraction in experiment 

 14 occurred during the first minute of the trial : 

 the scores during the first minute differed by 57 

 percent; the total scores differed by only 25 per- 

 cent. This fact suggests that light or dark adapta- 

 tion affects the initial attraction to light more than 

 it affects the tendency for the light to liold the fish. 



Although the total scores at high temperature in 

 experiment 15 did not differ significantly between 

 light- and dark-adapted herring, tlie first minute 

 scores were somewhat (though not significantly) 

 higher for light-adapted herring. Experiment 16 



FACfTORS INFLUENCING ATTRACTION OF ATLANTIC HERRING TO ARTIFICIAL LIGHTS 



81 



