514 



Fishery Bulletin 103(3) 



1/15/2001 



1/29/2001 



2/12/2001 



2/26/2001 



3/12/2001 



3/26/2001 



Figure 9 



Comparison of relative abundance of Pacific cod (Gadus maerocephalus) in 

 portions of Steller sea lion lEumetopias jubatus) critical habitat from 15 Janu- 

 ary-24 March 2001 based on 1) no fishing model: the proportion of the maxi- 

 mum biomass (on 15 February) in critical habitat each day; 2) the fishing 

 model: subtracting catch per day from 15 January-24 March 2001 in high 

 and low sampling-effort areas from the no fishing model (total of 12,800 t); 3) 

 longline fishery catch-per-unit-of-effort (CPUE) index of abundance from the 

 high sampling-effort area, 13 February to 24 March (assigned a value of 1 on 

 13 February); and 4) trawl fishery (20% threshold) CPUE index of abundance 

 from the high sampling-effort area, 6 February to 24 March (assigned a value 

 of 1 on 6 February). 



surveys of the entire Bering Sea shelf conducted in 

 summer (Thompson and Dorn, 2002). This difference 

 stems from highly domed-shaped selectivity-at-length 

 schedules for the summer surveys and most fishery 

 catches of cod (Thompson and Dorn, 2002). As a conse- 

 quence, the model "assumes" that fewer cod are caught 

 in proportion to their actual abundance at lengths 

 greater than 45 cm for the survey catch and 80 cm 

 for the fishery catch. However, it is unclear how large 

 cod avoid capture during surveys or by longline, pot, 

 and trawl fishery gear as implied by the dome-shaped 

 selectivity-at-length schedules. 



A seasonal model of Pacific cod movement patterns 

 into and out of Steller sea lion critical habitat (Fig. 2) 

 indicates that relative Pacific cod biomass inside criti- 

 cal habitat is highest in February, then drops 13% in 

 March and 44% by April. If these values are assigned 

 to the middle of each month and daily values are ex- 

 trapolated linearly, the relative change from 15 Febru- 

 ary through 24 March is 23% (Fig. 9). Fishery indices 

 of abundance in the HSE area in January and Febru- 

 ary are consistent with this seasonal pattern, with 

 both trawl and longline CPUEs increasing from Janu- 

 ary to February. According to Figure 2 and the 2001 

 age 3+ biomass estimate (Thompson and Dorn, 2002), 

 catches through 24 March within the entire survey area 

 (12,806 t) represented only 1% of the BSAI stock and 



should have reduced the relative biomass of cod within 

 critical habitat by only an additional 2%. Thus, the 

 total reduction in relative cod biomass within critical 

 habitat from mid-February through late March after 

 accounting for fishing and emigration should have been 

 25% (Fig. 9). Longline and trawl fishery CPUE data 

 in the HSE area provide an independent estimate of 

 relative cod biomass. Both indices indicate that the re- 

 duction in relative cod biomass within the HSE survey 

 area through 24 March was 71-46% greater than that 

 predicted by the model. 



Catches and biomass estimates of Pacific cod for dif- 

 ferent time periods and areas can be used to compute 

 harvest indices (catch divided by observed biomass). 

 For instance, the harvest index within the entire sur- 

 vey area (based on the catch from 1 January through 

 24 March and the survey biomass estimate) was 26% 

 (12,806 or-=-49,032). If the focus is narrowed to only the 

 HSE survey area through 24 March, the harvest index 

 was 37% (11,631 or^-31,312). However, both the fish and 

 the fishery were concentrated within the HSE area. The 

 eastern two-thirds of the HSE survey area had survey 

 and fishery-derived biomass estimates of 23,418 t and 

 -14,500 t, respectively. With the area of fishery effort 

 more precisely defined, local harvest indices increase 

 even further, ranging from 50% (11,631 or-23,329) to 

 80% (11,631 or-r 14,500). 



