by means of his dip-net. Great numbers of herring 

 are attracted by the light and it is not uncommon 

 for fifteen or twenty barrels to be taken in a few 

 hours." 



The variability of the herring's response to light 

 is a characteristic feature and has been noted by 

 many investigators. Blaxter and Holliday (1963) 

 stated ". . . . Such i-eactions will vary widely de- 

 pending on the environment and the physiological 

 state and age of the fish as well as on the type of 

 stimulus itself." Other authors have demonstrated 

 how many factors, external and internal, can vary 

 the response of herring and other species to artifi- 

 cial lights. Kurc (in press) and others have pointed 

 out how the thermocline may prevent fish from 

 rising to a light, or may hold them in the surface 

 water so that they can be more readily attracted. 

 Strong ambient light (e.g. moonlight) may reduce 

 the effectiveness of the attracting light (Kurc, in 

 press; Strom, in press). Woodhead (19.56) showed 

 that starvation reversed the normal negative 

 phototaxis of tlie minnow Plioxlrms. Andrews 

 (1946) found that the attraction to light of the 

 white sucker {C atastomnis) decreased with increas- 

 ing temperature. Sudden changes in illumination 

 may cause the fish to disperse instead of attracting 

 them (Strom, in press; Gauthier, in press). 



Although routine, uncritical use of lights to at- 

 tract fish may sometimes be successful, far greater 

 effectiveness might be a/chieved by a better imder- 

 standing of the underlying behavior of the fish 

 and its response to lights. Moreover, lights might 

 be of definite value in some circumstances where 

 they are not now used. In Maine the use of lights 

 for catching herring is generally illegal because 

 many fishei-men believe that the lights tend to dis- 

 pei-se the herring rather than attract thean (Scat- 

 tergood and Tibbo, 1959). This restriction appears 

 to be an instance where profitable use of lights has 

 been discouraged because the tmderlying beliavior 

 of the fish has been inadequately understood. 



This paper is an attempt to explain some of the 

 biological and other factors that are conducive to 

 the attraction of herring by light. 



METHODS 



Tlie fish iLsed in the experiments were immature 

 Atlantic herring of age groups O (brit), I, and 

 II, which are processed as Maine sardines. They 

 were 75 to 200 mm. in total length and were taken 



from conunei'cial catches near Boothbay Harbor, 

 Maine. The fish were held under the prevailing 

 seasonal conditions of salinity, temperature, and 

 dissolved oxygen in large tanks provided with iian- 

 nLng sea water and were fed daily a mixture of 

 ground trout food and canned cat food. 



The experiments were conducted in a separate 

 tank. This tank was fiberglass, 5.5 m. long, 0.4 m. 

 deep, and 0.3 m. wide. An incandescent lamp sus- 

 pended in a glass cylinder at each end of the tank 

 provided the attracting illumination; by raising 

 or lowering the lamp bulb in the cylinder, the light 

 source could be located above or below the surface. 

 Sea water entered the left-hand end of the tank 

 and drained off at the right. A slight drift (less 

 than 4 cm. per minute) toward ih^ right resulted. 

 Tliis and other sources of left-right bias were c-om- 

 liensated by periodically alternating the location 

 of the light source between the right and left, ends 

 of the tank with a double thi-ow switch. For tem- 

 peratures above the seasonal sea-water tempera- 

 ture, the incoming water was lieated. An air Ixib- 

 bler near the point of entrance and another near 

 the center of the tank provided sufficient mixing 

 so that temperature differences within the tank 

 did not exceed 1° C. Cooling pipes, carrying a 

 chilled ethylene glycol-water mixture, located 

 along the walls of the tank provided uniform re- 

 frigeration when below-seasonal temj^eratures 

 were required. 



Variations in dissolved oxygen were achieved 

 by recirculating the water through a tank of pure 

 oxygen under pressure (1.5-2.0 atmospheres) or 

 through a vacuum. These devices provided a range 

 of 50 percent to 250 percent oxygen saturation in 

 the experimental tank. 



The intensity of the attracting lights was varied 

 by using light bulbs of different wattages or by 

 varying the supply voltage. The light gradient 

 for each intensity and source position (fig. 1) was 

 measured with a photovoltaic light meter* having 

 a waterproof housing for the sensitive element. 

 This element was held in a plane normal to the 

 directiion of the light rnys in the water. The meter 

 was factory-calibrated for a spectral response cor- 

 responding to that of the human eye. This response 

 is not identical to that given for herring (Blaxter, 

 1964), but is very similar: the difference was so 

 small that special calibration of the instrument 

 did not seem warranted. 



74 



U.S. FISH AND WILDLIFE SERVICE 



