490 



Fishery Bulletin 102(3) 



Sampling dates 



Sampling was conducted biannually, in April and Sep- 

 tember, from 1992 to 1997 and annually, in September, 

 from 1998 to 2000 (Table 1). The spring and fall sam- 

 pling periods were selected to coincide with crab life 

 history events and to avoid sampling during commercial 

 fishery operations. April sampling was scheduled to 

 occur before larval hatching in May- June (Shirley et 

 al., 1987) and before the summer commercial fishing 

 season from 15 June to 15 August. September sampling 

 began after the end of the fishing season (15 August) 

 and ended before the beginning of the winter harvest 

 (1 October to 30 November). 



During 1992, the study sites were sampled with pots 

 (referred to as "pot sampling") and by divers (referred to 

 as "dive sampling") concurrently (Table 1). In 1993 and 

 1994, sampling was conducted on nearby study sites 

 and the dive sampling usually one day ahead of the 

 pot sampling. For logistical reasons, starting in 1995. 

 we separated the pot sampling and the dive-transect 

 sampling into two separate research cruises. The pot 

 sampling was conducted on the first cruise and the 

 dive sampling occurred on the second cruise; pot and 

 dive sampling were separated at each location by 2 to 

 12 days. 



Sampling with pots 



Crabs were sampled with commercial crab pots (0.91 m 

 in diameter, 0.36 m tall, with 5-cm wire mesh). Escape 

 rings were sealed with webbing on each pot to retain 

 smaller crabs. Pots were baited with hanging bait com- 

 prising salmon, cod, or halibut (depending on availabil- 

 il nd bait jars that were filled with chopped herring 

 and squid. We found that cod was predictably available; 

 therefore from 1996 on, we consistently used cod for 

 hanging bait. Pots were soaked for 24 hours. 



Within each study site, we set 25 pots in shallow 

 water (0-9 m) and 25 pots in deep water (10-25 m). 

 Each day we set 50 pots in one of the study sites and 

 retrieved the 50 pots that had been set the previous day 

 at one of the other study sites. The pots were set along 

 strings parallel to shore at intervals of approximately 

 100 m. Within each study area, the strings of pots were 

 located in prime Dungeness crab habitat determined 

 by a local fisherman. We placed the pots at the same 

 locations during subsequent sampling events by using 

 a GPS (Rockwell PLGR+) with an accuracy of ±3 m. We 

 estimate that the pots were set within 20 meters from 

 the original waypoints. Water depth (standardized to 

 mean lower low water), set and retrieval time, and GPS 

 location were recorded for each pot. Water temperature 

 and salinity profiles were measured at each study site 

 during each sampling period with a SEABIRD SBE-19 

 Profiler. 



As the pots were retrieved, we counted and identified 

 all organisms. For all Dungeness crabs we recorded the 

 sex, carapace width, shell condition, and damage to 

 appendages. For female crabs we also recorded repro- 

 ductive status. Carapace width was measured to the 

 nearest millimeter immediately anterior to the 10th 

 anterolateral spine with vernier calipers (Shirley and 

 Shirley, 1988; Shirley et al.. 1996). All organisms were 

 returned to the water at the location where they were 

 caught. A potential problem with returning the crabs 

 to the water near the site of capture is the possibility 

 that crabs could be resampled in subsequent pots, which 

 would bias the catch per unit of effort. Beginning in 

 April, 1995, all crabs collected in the South Beardslee 

 Islands and Berg Bay were tagged with a sequentially 

 numbered, double-T Floy tag (Floy Tag and Manufactur- 

 ing Company, Seattle, WA) inserted along the postero- 

 lateral margin of the epimeral suture. Tags placed in 

 this location are retained through ecdysis (Smith and 

 Jamieson, 1989). Of the 5226 crabs tagged, only a single 



