880 



Fishery Bulletin 97(4), 1999 



spits occurred in less than 5 seconds, and 95% oc- 

 curred in less than 25 seconds. Once hooked, and 

 after initial rushing, the halibut lay on the bottom, 

 resting, then went into a pattern of rushing, resting, 

 rushing, etc. There appeared to be a pattern of de- 

 creasing duration of rushing following subsequent 

 rest periods. The average times for resting and rush- 

 ing after the initial hooking were 3 min 49 s, and 31 s, 

 respectively (Table 2). 



Hooking success for other species 



The range of approach and interaction behaviors for 

 species other than halibut was not documented be- 

 yond noting bitings and subsequent hooking. Other 

 species that bit at hooks included canary rockfish 

 iSebastes pinniger), yelloweye rockfish iS. ruberrim us ), 

 quillback rockfish iS. maliger), raffish (Hydrolagus 

 colliei), and lingcod (Ophiodon elongatus). In general, 

 hooking success for the major rockfish species was 

 around 5-6% (Table 5). These fish were comparable 

 in average size (weight) to those caught as inciden- 

 tal catch in the directed halibut fishery. 



Discussion 



The feeding behavior of halibut may be classified into 

 three phases; arousal, search, and bait attack and 

 food ingestion. 



Arousal 



In most cases, arousal to the presence of food occurs 

 at a distance, and bait odor carried by bottom cur- 

 rents attracts fish from well beyond the limited view- 

 ing distance of the present experiment. Although 

 undoubtedly involving detection of an odor plume 

 (Atema, 1980), the present experiment did not investi- 

 gate the initial arousal phase of feeding behavior 



However, a number of observations in relation to food 

 location and uptake were made. 



Search 



It is clear that halibut use orientation to bottom cur- ' 

 rent to locate food, most approaching upstream when 

 a bottom current was present. This rheotactic orien- 

 tation is used by many fish for detection and loca- 

 tion of prey (Atema, 1980; L0kkeborg et al., 1989; 

 Lokkeborg, 1998). Once fish are aroused to the pres- 

 ence of a prey item by scent, they orient into a cur- 

 rent to locate the food. 



The role of vision in food location by Pacific hali- 

 but was less clear. Behaviors directed towards the 

 gear were observed in some cases immediately after 

 the gear reached the seafloor. These were likely re- 

 sponses to visual clues. However, the greatest num- 

 ber of appearances appeared to be motivated by scent 

 carried by the current. Although earlier studies have 

 shown that halibut certainly prey on a number of 

 pelagic or semipelagic species where vision must play 

 an important role in prey recognition and capture 

 (Best and St-Pierre, 1986), it probably played a lim- 

 ited role in behavior toward the model setline. 



Prebiting behavior and attack rate 



Fish vary widely in their reliance on sight, smell, 

 and touch in deciding whether to accept or reject food 

 once it has been located ( Lokkeborg, 1994 ). Both loop- 

 ing and lying behaviors were common in the observed 

 halibut, and in only a few cases was a complete bite 

 initiated without some preliminary bait interaction 

 such as looping or lying. Looping behavior could be a 

 test of an odor plume by the fish, assuring that the 

 fish is in an area of high-scent concentration. Many 

 halibut lay near the bait prior to initiating a bait 

 attack, most either directly downstream or to one 

 side of the bait in relation to the current. This behav- 

 ioral response may reinforce olfactory clues that led 

 the fish to the bait, or possibly a restrained response to 

 the baited hook as a novel prey item (Lokkeborg, 1990). 

 Rejection of the bait at this point was often followed 

 by looping, behavior that often led to further bait 

 interaction. The role of mechanoreception was not 

 demonstrated. In only a few instances did halibut 

 exhibit incomplete bites, and these generally did not 

 result in the fish then leaving the area of the gear. 

 Most bites were associated with extremely active be- 

 havior, e.g. a fish beginning to rush simultaneously 

 with biting the bait. Only one bite that resulted in 

 hooking was not followed by a rush. 



In the simplest interpretation, the attack rate for 

 halibut was 43%. Fish length was not a significant 



