504 



Fishery Bulletin 103(3) 



1 o 



09 

 08 

 07 

 06 

 05 

 04 

 03 

 02 

 1 

 0.0 



Feb Mar Apr May Jun 



Aug Sep Oct Nov Dec 



Figure 2 



Proportion of maximum (in February) biomass of Pacific cod iGadus 

 macrocephalus) within Steller sea lion (Eumetopias jubatus) critical 

 habitat in the eastern Bering Sea by month (see Footnote 10 in the 

 general text). 



against which possible changes related to local fisheries 

 can be compared. The model results indicate that the 

 highest biomass in critical habitat (largely on the shelf 

 north of Unimak Island) occurs in February, declines to 

 about 10% of the peak in June, and then slowly rebuilds 

 through the summer and fall. Changes in the behavior 

 of Pacific cod immediately prior to or after spawning, 

 such as the formation of dense aggregations or the tem- 

 porary cessation of feeding, would affect catchability by 

 both trawl and fixed gears. However, abrupt changes in 

 catchability due to the formation of aggregations should 

 be evident within the time series of catch and effort 

 data, and changes in feeding habits would not affect 

 the catchability by trawl gear. 



Methods 



Bottom trawl survey 



Stations sampled during the bottom trawl survey were 

 selected by using a stratified random scheme. Two strata 

 were defined: one with a high and another with a low 

 degree of sampling effort, based on the expected distri- 

 bution and abundance of Pacific cod from fishery infor- 

 mation. In the nearshore or high sampling-effort stratum 

 (7765 km 2 ), 38 stations were sampled, whereas 19 sta- 

 tions were sampled in the larger (12,112 km' 2 ), offshore 

 low sampling-effort stratum (Fig. 1). All survey tows 

 were conducted during daylight hours from 16 Febru- 

 ary to 1 March 2001 aboard the FV Northwest Explorer 

 and the FV Ocean Harvester. The 49-m FV Northwest 

 Explorer was equipped with two 1800-hp engines, and 

 the 33-m FV Ocean Harvester had a single 1250-hp 

 engine. Both vessels were house-forward trawlers that 

 had stern ramps, multiple net storage reels, and paired 



hydraulic trawl winches with 1280-2190 m of 2.54-cm 

 diameter steel cable. Each vessel carried a full comple- 

 ment of navigation and fishing electronics, including 

 global positioning systems (GPS), video position plotters, 

 radars, and depth sounders. 



A Poly-Nor'eastern high-opening bottom trawl rigged 

 with roller gear was used to sample the groundfish 

 community at each selected location. The trawl net 

 was constructed of 12.7-cm stretched-mesh polyethylene 

 web and had a 3.2-cm stretched-mesh nylon liner in 

 the codend. Accessory gear for the Nor'eastern trawl 

 included three 54.9 m, 1.6 cm diameter galvanized wire 

 rope bridles, and 1.8 x 2.7 m steel V-doors weighing ap- 

 proximately 850 kg each. 



Biomass (S) estimates for each stratum surveyed 

 were computed by multiplying the average CPUE (in 

 units of kg/km 2 ) for all hauls (n) in a stratum by its 

 area (A). Haul CPUE was calculated as the weight of 

 cod caught (kg) divided by the area swept (a), which 

 was the length of the tow multiplied by the average net 

 width determined by sonic mensuration equipment: 



kg 



■xA. 



B 



Confidence bounds on stratum biomass estimates were 

 computed from the standard deviation of the haul 

 CPUEs. For haul CPUEs we assumed a catchability 11 

 of 1 for Pacific cod (all cod within the area swept by 



Note that catchability within the survey biomass estima- 

 tion procedure has a different literal definition than in the 

 Leslie equation. 



