Kaimmer: Hooking behavior of Hippog/ossus stenolepis 



877 



function of the matrix was to show possible pat- 

 terns or combinations of behaviors that occurred 

 more or less frequently than expected if all pos- 

 sible behavior transitions had the same probabil- 

 ity of occurring. Statistical analysis of this matrix 

 is complicated by the fact that not all transitions 

 were possible (a spitting may only follow after bit- 

 ing or rushing). The matrix was therefore visu- 

 ally inspected for large differences, as recom- 

 mended by Slater ( 1973 ). 



Prebiting behavior and attack rate 



There were 293 transitions observed for halibut, 

 not including behaviors that followed initial com- 

 plete bitings. The number of prebiting transitions 

 per individual fish ranged from one to fifteen, with 

 78% of the fish making three or less transitions, 

 and over 90% making five or less. 



Looping and lying near bait were the most com- 

 mon behaviors following initial observation. Less 

 commonly observed were departures, biting, and 

 incomplete biting. Looping behavior was followed 

 about equally by lying near bait, biting, and depar- 

 ture. Lying near bait was most often followed by 

 either looping or biting, and less often by departure 

 or an incomplete biting. For 71 occurrences of lying 

 behavior, 8 either had no obvious hook orientation 

 or occurred during slack current, 3 fish lay upstream 

 of bait, 26 lay to the side of bait, and 34 lay down- 

 stream from bait. The distance from bait for fish ly- 

 ing beside or downcurrent averaged 13.2 cm (SD=8.4 

 cm) and 14.6 cm (SD=9.0 cm), respectively. Incom- 

 plete biting was most often followed by lying near 

 bait, and less often by departing or looping. 



The transitions in Table 3 do not consider preced- 

 ing behaviors, or how many previous behaviors a 

 particular fish had completed. As each of the origi- 

 nal 129 halibut was observed through the first four 

 behavior transitions up to either biting or depart- 

 ing, there was a tendency for fish to stay in interac- 

 tion with baits. Lying, incomplete and complete bit- 

 ing, and grazing are all behaviors that are direct in- 

 teractions with either the bait or food present around 

 a baited hook. In the first transition (n = 129), fish 

 were most likely to interact with bait (43%), loop 

 (34%), or depart (23%). About one-third of the bait 

 interactions in the first transition were bites. In the 

 second transition (n=78), fish were most likely to in- 

 teract with the bait (57% ) or depart (40% ). Only 12% 

 looped. Again, about one-third of the bait interac- 

 tions were bites. By the third transition {n=28), fish 

 were much more likely to continue bait interactions 

 (68%) than they were to loop (25%) or depart (7%). 

 About one-quarter of bait interactions were bitings. 



E 



^ Biting 

 D Not biting 



2 



^^Z^ 



Upstream Side Downstream Slack 



Approach direction relative to currei 



Figure 3 



Numbers of halibut biting and not biting, by approach direc- 

 tion in relation to bottom current. 



In the fourth transition (n=21), 62% continued with 

 bait interactions, almost half of which were bitings, 

 21% departures, and 19% loopings. By the end of the 

 fourth behavior transition, out of the original 129 

 halibut observed, approximately half (66 fish) had 

 departed, whereas just over a third (47) had made 

 complete bites. In each of the transition levels, 43- 

 68% of behaviors were directed toward baits. 



Elapsed time between appearance and the first 

 observed bite for all halibut ranged from 1 s to 7 min 

 54 s (mean=34 s, SD=1 min 5 s). Ninety-five percent 



