Taggart et al.: Estimating abundance of Cancer magister 



489 



58°30'N. 



^s 20' 



t r 



135°50' 135 4n\\ 



Figure 1 



Map of study area showing eight study sites in or near Glacier Bay in southeastern 

 Alaska. 



history periods such as mating and molting (Kruse, 

 1993; Orensanz et al., 1998). Pre- and postseason stock 

 assessment surveys using crab pots were initiated in 

 southeastern Alaska in 2000 (Rumble and Bishop, 

 2002). The purpose of the latter management strategy 

 is to assess the abundance of legal-size males before the 

 fishing season, to estimate harvest rates, to define the 

 timing of male and female mating and molting and to 

 determine growth rate by tagging crabs. 



The usefulness of surveys with pots for Dungeness 

 crab population assessment, however, depends on the 

 accuracy of these surveys in measuring population pa- 

 rameters. Factors that can bias catch per unit of effort 

 (CPUE) and size-frequency estimates for Dungeness 

 crabs are pot soak-time (Miller, 1974; High, 1976; Got- 

 shall, 1978; Smith and Jamieson, 1989); freshness of 

 bait (High, 1976; Smith and Jamieson, 1989); pot design 

 (Miller, 1974; High, 1976; Smith and Jamieson, 1989); 

 and agonistic interactions among conspecifics inside 

 and at the entrance of pots (Caddy, 1979; Smith and 

 Jamieson, 1989). Smith and Jamieson (1989) developed 

 a standardized model to compensate for the effect of 

 agonistic interactions, age of bait, and escapement. 

 They also concluded that researchers could minimize 

 these biases by measuring CPUE with standardized 

 surveys with short soak times. These studies measured 

 sampling bias with pots by comparing catch in pots 

 among various experimental treatments. Opportunities 

 for comparing surveys with pots to direct measures of 

 abundance are rare. In our study, we compared the bias 

 and power of CPUE estimates from surveys with pots to 



independent measures of abundance conducted by scuba 

 divers on benthic dive transects. 



Methods 



Study area 



The study area included eight sites in southeastern 

 Alaska, near Glacier Bay: North Beardslee Islands 

 (58°33'N 135°54'W), South Beardslee Islands (58°30'N 

 135°55'W), Berg Bay (58°31'N 136°13'W), Bartlett Cove 

 (58°27'N 135°53'W), Gustavus Flats (58°23'N 135°43'W), 

 Secret Bay (58°29'N 135°56'W), inner Dundas Bay 

 (58°27'N 136°31'W), and outer Dundas Bay (58°21'N 

 136°18'W) (Fig. 1). All study sites were located within 

 Glacier Bay National Park and Preserve, with the excep- 

 tion of Gustavus Flats, which was located adjacent to the 

 Park boundary in Icy Strait. 



Glacier Bay is a large (1312 km 2 ) glacial fjord system 

 with high sedimentation rates of clay-silt particles from 

 streams and tidewater glaciers (Cowan et al., 1988). 

 The primarily unconsolidated rocky coastline is highly 

 convoluted with numerous small bays. Dungeness crabs 

 can be found throughout Glacier Bay; however the ma- 

 jority of the population are found in the lower 40 km of 

 the estuary where our sites were located (Taggart et al., 

 2003 ). The shallow water in and around our study sites 

 was primarily characterized by mud bottom, but sand, 

 pebble, cobble, and shell substrates were also common 

 (Scheding et al., 2001). 



