Taggart et al.: Estimating abundance of Cancer magister 



493 



transects by increasing the study duration or increasing 

 the amount of change in the population that the study 

 is attempting to detect (Fig. 6). Although pots had more 

 power than dive transects, there was only slightly more 

 power to detect change in abundance of male crabs 

 versus nonovigerous females (Fig. 7). 



Discussion 



For male Dungeness crabs, the density estimates from 

 the dive transects showed no difference between April 

 and September (Fig. 2B). The male CPUE estimates 

 from pots, however, were consistently lower in April than 

 in the following September (Fig. 2A). Because feeding 

 rates of Dungeness crabs are correlated with tempera- 

 ture (Kondzela and Shirley, 1993), we thought that tem- 

 perature was likely to explain the differences in CPUE 

 between April and September. We found, however, that 

 season had a larger effect than temperature (Table 2). 

 This result suggests that seasonal factors other than 

 temperature influence catchability. Stone and O'Clair 

 (2001) followed the seasonal movements of Dungeness 

 crabs in a glacial estuary in southeastern Alaska and 

 reported that mean movement of male crabs was lower 

 during the spring than in the late summer and fall. It 

 is possible that our spring sampling schedule coincided 

 with low male activity and male crabs were less likely to 

 encounter a bait plume and be attracted to a pot. These 

 results indicate that if pots are used for sampling, late 

 summer and early fall is the time of year to conduct 

 population assessment surveys of male crabs. Similar 

 seasonal differences in CPUE have also been described 

 for edible crabs (Cancer pagurus) and American lobsters 

 (Homarus americanus) (Bennett, 1974). These data dem- 

 onstrate the importance of controlling for season when 

 comparing CPUE among years or sites. 



The proportion of large crabs caught in pots increased 

 with longer soak time for Dungeness crabs in British 

 Columbia (Smith and Jamieson, 1989) and red king 

 crabs in Britstol Bay, Alaska (Pengilly and Tracy, 

 1998). We found no bias when we measured the legal- 

 size proportion of the male population caught in pots 

 and compared it to the proportion sampled on dives 



1 



0.8 -I 



0.6 



0.4 -I 



0.2 





 1 



0.8 



0.4- 

 0.2- 



B 



April P=0.17 

 Sept. P=0.49 



' A 



0.6 9 o 



o 



i>4. 



O " Oft /A 

 *«0«' *A 



0°A^ 



a 



* A 



A AA 



o 



O April 



A September 



0- 



i-i c 



0.8 

 0.6 

 0.4 

 0.2 



0* 



April P<0.0001 

 Sept. P<0.0001 



O 



o 



cr 

 o 



0.2 0.4 0.6 0.8 1 

 Percent crabs on transects 



Figure 3 



The percentage of (A) female Dungeness crabs 

 (C. magister). (B) nonovigerous female crabs, 

 and (Ci ovigerous female crabs estimated from 

 pots and from dive transects. The dashed line 

 in each graph has a slope of 1: thus half of the 

 data points should be above and half should be 

 below the dashed line if percentage estimates 

 for dives and pots are unbiased. Pot and dive 

 transect data for each sex class and season 

 were compared with a paired sign test and 

 P-values are reported. 



(Fig. 4). We expect, however, that the bias observed in 

 British Columbia and Bristol Bay would occur for our 

 study sites if the soak time of pots were increased. 



