Fritz and Brown: Interactions between the Pacific cod fishery and Steller sea lions 



509 



other species were targeted. On the other hand, the use 

 of a high threshold (such as 60%) might exclude hauls 

 where Pacific cod was the target species. Therefore, a 

 range of trawl target definitions from 20% to 60% was 

 used. The cod trawl fleet distribution shown in Figure 

 4 was defined by the 20% threshold. If the 40% or 60% 

 thresholds are used, most of the cod trawl effort shown 

 in the HSE area remains, whereas some of the effort in 

 the eastern portions of the AE of the survey area is not 

 coded as the effort of a cod-target fishery. 



Distribution of Pacific cod catch Cod catches accu- 

 mulated differently in the three primary areas fished 

 (Fig. 5). In the HSE area, cod catches rose steadily from 

 1 January through early April, and totaled approxi- 

 mately 13,000 t. There was a brief increase in the rate 

 of cod catch in mid-April, but by approximately 20 April, 

 the cod fishery in the HSE area had essentially finished 

 with a catch total of 17,875 t. In the AE of the survey 

 area, cod catches accumulated steadily from 1 Janu- 

 ary through 2 March, and totaled 6340 t. There was a 

 brief increase in catch rates for 6 days from 25 through 

 30 March, after which the cod fishery in the AE of the 

 survey area was finished with a catch total of 7691 1. In 

 the AS of the survey area, there was little cod fishing 

 effort prior to 22 February, and it lasted only through 

 27 March, by which time almost 3500 t had been caught; 

 catches through 30 April from the AS of the survey area 

 totaled 3724 t. There was very little cod fishery effort 

 in the LSE area (Table 3), and only 1200 t of cod were 

 caught (principally as bycatch in other fisheries) through 

 30 April 2001. 



The longline fleet began fishing for Pacific cod in both 

 the HSE area and AE of the survey area on 1 January 

 (Fig. 5). In the HSE area, daily average longline CPUE 

 (t cod per 1000 hooks per day) remained relatively low 

 and steady, ranging from 0.3-0.7 through January. The 

 longline fleet left the HSE area for approximately two 

 weeks, resuming effort again on 13 February and con- 

 tinuing through 6 March. Longline CPUEs were gener- 

 ally higher in late February than they were in January, 

 ranging from approximately 0.7 to 1.2. The longline 

 fleet again returned to the HSE area on 19-24 March, 

 but daily average CPUEs were <0.5. There was sporadic 

 longline fishing for cod in the HSE area through April, 

 and CPUEs ranged from 0.3 to 1.0. In the AE of the 

 survey area, the longline fleet fished continuously from 

 1 January through 2 March, and daily average CPUE 

 declined from a range of 0.7-1.0 on 1-7 January to a 

 range of 0.3-0.5 on 24 February-2 March. 



The trawl fishery for cod began on 20 January in both 

 the HSE area and AE of the survey area (Fig. 5). In the 

 HSE area, trawl CPUE (t cod per hour trawled per day) 

 increased from a range of 0.7-1.4 on 20-27 January to a 

 range of 1.3-2.5 on 6-15 February. From 16 February- 

 1 March, trawl CPUEs were slightly lower, ranging from 

 0.8 to 2.0, after which they declined further, ranging 

 only from 0.5 to 1.3 from 2-24 March. On 26 March, 

 the average CPUE increased substantially to over 12 

 but quickly declined to less than 1.0 by 1 April. This 



was followed by another short-lived increase in CPUE 

 on 11 April, after which daily average CPUEs remained 

 below 1.0 through April. In the AE of the survey area, 

 CPUEs were highly variable (between 0.4 and 2.3) and 

 there was little observable trend between 20 January 

 and early March. On 25 March, however, average CPUE 

 increased to over 4 and ranged between 0.4 and 3.9 

 through 2 April, after which there was only sporadic 

 effort and daily average CPUEs were less than 1. 



The pot fishery for cod began on 22 February south 

 of the survey area and on 24 February in the HSE area 

 (Fig. 5). In the AS of the survey, pot CPUE (t cod per 

 20 pots per day) decreased from a range of 0.3-1.0 from 

 22 February-1 March, to a range of 0.2-0.5 on 8-17 

 March. However, on 18 March, pot CPUE increased 

 to 1.1, and remained between 0.5 and 0.8 through 22 

 March, after which it quickly declined to very low lev- 

 els. In the HSE area, pot CPUE ranged between 0.7 and 

 1.7 from 24 February to 23 March. However, on 24-25 

 March, CPUE was greater than 2. Pot cod fishing oc- 

 curred on only three more days through the end of April 

 in the HSE area: on 27 March, 6 April, and 12 April. 

 Although daily average CPUEs on the last two days 

 were the highest recorded in the pot fishery in 2001, 

 observed catches on these days totaled only 4 and 5 t 

 of cod, respectively. 



Leslie depletion analyses Leslie depletion analyses 

 were conducted on four sets of Pacific cod fishery data 

 collected in the HSE area and on two sets of data col- 

 lected in the AE of the survey area (Table 5). In the 

 HSE area, longline fishery data collected prior to 13 

 February and trawl fishery data collected prior to 6 

 February were excluded from the analyses because 

 CPUE data indicated that fish were immigrating 

 into the area in January in preparation for spawning 

 (Fig. 5). It is unlikely that the increase in CPUE was 

 due to a change in catchability because the increase 

 was evident whether bait was used (pots and longlines) 

 or not (trawls). Data indicating an increase in the 

 abundance of cod north of Unimak Island in January 

 and a peak in February were in agreement with a 

 generalized model of cod abundance in Steller sea lion 

 critical habitat in the eastern Bering Sea (Fig. 2) and 

 seasonal cod movements from tagging data (Shimada 

 and Kimura, 1994). The time series was truncated at 

 24 March because of the evidence within the fisheries 

 data (increase in CPUE) that another group of cod had 

 immigrated to the HSE area and AE of the survey 

 area in late March or that catchability had increased 

 substantially (Fig. 5). In addition, daily average CPUEs 

 from hauls that had at least 20%, 40%, and 60% Pacific 

 cod by weight were regressed against cumulative catch 

 to see what effect the target definition might have on 

 the regression results. 



All Leslie regressions with longline or trawl fish- 

 ery data from the HSE area were highly significant 

 (P<0.000001; Table 5 and Fig. 6). Coefficients of de- 

 termination (r 2 ) for the longline and the trawl-20% 

 data were both greater than 0.6. Regression coefficients 



