Figure 8. — Impoundment for herring, showing quad- 

 rants used to record observations. 



was only one period (30 to 45 minutes) during 

 which no fish were in the upper layer; this 

 distribution occurred about an hour before sun- 

 set (2000). Similarly, there was only one 

 period during which no fish were located in the 

 lower layer. This period began one-half hour 

 after sunset and continued for 45 to 60 minutes, 

 after which herring were dispersed throughout 

 the holding pen. 



During daylight herring were more numerous 

 in the upper area of quadrants A and B than 

 in C and D. This difference probably is ex- 

 plained by the uneven di.stribution of light. 

 Quadrants C and D were located closest to the 

 vessel float which was used to anchor the hold- 

 ing net. The upper areas of these quadrants 

 were shaded by the float and were avoided by 

 the fish. The distribution of herring in the lower 

 layers of A-B and C-D was relatively uniform. 

 After dark, fish were equally dispersed in the 

 upper and lower layers of the pen. 



On the day of the eclipse, fish were distrib- 

 uted in the upper layers of quadrants A and B 

 and in the lower layers of quadrants A, B, C, 

 and D. Fi.sh were absent from the upper layer 

 of quadrant D until sunset. No fish were ob- 

 served in the upper layer of quadrant C from 

 the start of observations at 1600 until 1730 — 

 15 minutes before totality. Fish were active 

 in this quadrant from 17.S0 until 1800; were 

 absent at 1815; and reappeared from 1830 to 

 1900. This movement of fi.sh into quadrant C 

 at the approach of totality was coupled with a 



change in behavior of herring in the other 

 quadrants. Some fish began to school and 

 moved to the surface of the water. This was 

 in contrast to the preeclipse behavior of gen- 

 eral dispersion without movement at the sur- 

 face, and apparently was in response to the 

 reduced light during the eclipse. The response 

 was not strong, and not all fish reacted to the 

 change. The data on subsurface illumination 

 (mentioned above) suggests that the duration 

 of lowered light intensity at totality was not 

 enough to stimulate a stronger response. This 

 conclusion is supported by the observations at 

 sunset, when the decrease in light approximat- 

 ed that of the eclipse, but for a longer period, 

 and elicited a stronger schooling response from 

 the herring. 



ECHO SOUNDINGS 



Echo sounders aboard the two ve.ssels were 

 run continuously during the study, but none 

 of the records showed any change or movement 

 of organisms during the eclipse. Echo tracings 

 in Boothbay Harbor documented considerable 

 activity in the early evening on July 19 and 20 

 (fig. 9). On July 19, activity was first detected 

 at 1854 when minor peaks and streaks extended 

 up from the bottom. By 1955, some of these 

 streaks and dots were no longer in contact with 

 the bottom ; others remained in contact but 

 were extended and more pronounced. By 2100, 

 only a few of these marks were in contact with 

 the bottom ; the rest were scattered from the 

 bottom to the surface. This phenomenon was 

 also recorded on the evening of July 20, though 

 somewhat later and less pronounced than on 

 the previous evening. 



None of the organisms taken in the plankton 

 nets was large enough to account for the mark- 

 ings observed on the recording paper, and the 

 scheduled collections did not allow time to 

 utilize other gear on July 19 or 20. On July 21, 

 a trawl was fished during late evening in the 

 same area. Echo soundings were similar to 

 those on July 19 and 20. Large catches of 

 jellyfish (Aurelia aurita) were taken in the 

 net; presumably they were the animals detect- 

 ed by the sounder, but the possibility exists 

 that the tracings were from herring or other 

 fish that escaped capture. In any case, the lack 



266 



U.S. FISH AND WILDLIFE SERVICE 



