FISHERY BULLETIN: VOL. 82. NO. 2 



activity in surface water at night compared with 

 daytime, and diel vertical ascent of salmon into 

 near-surface waters at night. The higher catches 

 in the upper 2 m of the gill net at night than day 

 lend support to the last possibility, but not to the 

 exclusion of the other possibilities. 



Most other authors favor vertical migration as 

 an explanation for diel peaks in gill net catches 

 (Taguchi 1963; Manzer 1964; Mishimaet al. 1966). 

 Birman (1964) noted visual avoidance of "sweep 

 nets" by day, but concluded that salmon migrate 

 into upper waters primarily as a response to verti- 

 cal movements of their zooplanktonic prey which 

 they feed on during periods of low light intensity, 

 chiefly before dawn. 



Swimming activity could also influence catch- 

 ability, but neither Ichihara et al. (1975) nor Ichi- 

 hara and Nakamura (1982) found large differences 

 in day-night swimming speeds of chum salmon 

 tagged with ultrasonic transmitters. 



The most interesting finding of our study is the 

 distinct diel change in composition of major prey. 

 Stomach contents of sockeye, pink, and coho salm- 

 on were comprised largely of euphausiids after 

 sunset and during the night (Fig. 4). The largest 

 number of full stomachs, usually containing only 

 fresh euphausiids, were also found during the 

 nighttime. These three species of salmon preyed 

 intensively on euphausiids at night, often to the 

 exclusion of other types of prey. 



This change to feeding on euphausiids was first 

 observed in the salmon caught during the time 

 that the 24-kHz sonic scattering layer ascended 

 into surface waters (Fig. 5). We assume that 

 euphausiids were an important component of this 

 scattering layer (see Suzuki and Ito 1967). 



A 1.8 m Isaacs-Kidd midwater trawl collection 

 (three mesh sizes: 70, 11, and 4 mm stretch) in the 

 upper 10 m at night at the 24-h gill net station 

 caught mainly salps and medusae, but 

 euphausiids were abundant (19g/l,000 m 3 ). 

 Euphausiids were also abundant in a 1.3 m ring 

 net (1.0 mm mesh) towed at the surface after sun- 

 set at this station. The most common euphausiids 

 caught were Euphausia pacifica and Thysanoessa 

 longipes, the same species common in salmon 

 stomachs. Euphausia pacifica were found to 

 undertake diel vertical migrations at the Cana- 

 dian Weather Station located at lat. 50°N, long. 

 145°W (Marlowe and Miller 1975). Frost and 

 McCrone (1974) also found evidence for diel verti- 

 cal migration of E. pacifica at this location but not 

 for T. longipes. The intense predation on euphau- 

 siids at night is therefore thought to be related to 



their dense concentration and increased vulner- 

 ability in surface waters after dark. 



Most of the studies of the diel periodicity or 

 chronology of feeding in salmon have been 

 juveniles in fresh or coastal waters (Godin 1981). In 

 general, these indicate diurnal or crepuscular 

 feeding patterns for juveniles of pink salmon (Ali 

 1959; LeBrasseur and Barner 1964; Bailey et al. 

 1975; Parker and Vanstone 1966; Parker 1969; 

 Godin 1981), chum salmon (Bailey et al. 1975; M. 

 C. Healey as cited in Godin 1981), sockeye (Narver 

 1970; McDonald 1973; Doble and Eggers 1978), and 

 coho salmon (Mundie 1971). Bailey et al. (1975) 

 concluded that pink and chum salmon fry did not 

 feed during cloudy moonless nights. Nighttime 

 feeding by sockeye apparently occurs during 

 moonlight but not on cloudy or moonless nights in 

 Babine Lake (Narver 1970). Experiments con- 

 ducted by Brett and Groot (1963) and Ali (1959) 

 indicated that juvenile pink salmon changed their 

 mode of capturing prey below 10° mc (meter can- 

 dle), an intensity where the change from photo- 

 topic to scotopic vision apparently occurs, and 

 their feeding activity decreased between inten- 

 sities of 10° to 10 4 mc and most ceased between 

 10~ 3 and 10 5 mc. Experiments by Bailey et al. 

 (1975) showed almost no feeding by pink salmon 

 fry at light Intensities below 10 1 mc. 



In our study, salmon fed intensively on 

 euphausiids at night under an obscured, overcast 

 sky. From the general data given by Brown (1952) 

 and Blaxter (1970) we estimated that the light 

 intensities on this night were between 10 3 and 

 about 10 " 5 mc. But, despite these low light inten- 

 sities, with attendant reduction in contrast of prey 

 and sighting range to prey (Eggers 1977; Anthony 

 1981), salmon were capable of actively feeding on 

 small, euphausiid-sized prey. At night, larger prey 

 such as squids and fishes are probably encoun- 

 tered less frequently than euphausiids and evade 

 capture more easily because of reduced sighting 

 and tracking ranges of salmon. Euphausiids may 

 not be as capable of active predator evasion and, 

 when abundant in near-surface aggregations at 

 night, are encountered frequently and actively 

 selected. Bioluminescense produced by 

 euphausiids may facilitate detection and capture 

 by salmon. Thus, escape responses and sighting 

 ranges at different light intensities may influence 

 the size and type of prey selected at different times 

 of a diel period. 



Machidori (1968) reported that the indices of 

 stomach:body weight of sockeye and chum salmon 

 caught in gill nets that fished different depths 



396 



