Volstad et al Estimating dredge catching efficiency for Callinectes sapidus 



411 



Accurate estimates of absolute abundance and 

 population characteristics over time provide a means 

 of ensuring a sustainable harvest of the Chesapeake 

 Bay blue crab stock. CPUE must be adjusted for the 

 dredge catching efficiency to estimate absolute abun- 

 dance from the survey data. Catching efficiency (i.e. 

 the fraction of crabs present in the path of the dredge 

 that is captured) can be estimated from removal expe- 

 riments (e.g. Seber, 1973; Ricker, 1975; Hilborn and 

 Walters, 1992). In such experiments, a closed popu- 

 lation is sampled repeatedly over a relatively short 

 time. An estimate of the catching efficiency is typi- 

 cally based on the slope of a linear regression of CPUE 

 on cumulative catch (Leslie and Davis 1939), or on 

 log-transformed CPUE and cumulative effort ( Delury, 

 1947). It is assumed that no emigration, immigration, 

 or natural mortality occurs during the experiment 

 and that all animals caught are not returned to the 

 population (Otis et al., 1978; Schnute, 1983). 



During the summer months, blue crabs are active 

 swimmers; therefore, an otter trawl is more effec- 

 tive for sampling. Estimates of absolute abundance, 

 however, are difficult to obtain during the summer. 

 First, trawling at random locations may be difficult 

 because of the presence of crab pots and trotlines 

 throughout Chesapeake Bay. Second, the catching 

 efficiency of trawls is difficult to estimate because it 

 is affected by the swimming behavior of blue crabs. 

 The key assumption for estimating catching effici- 

 ency of a closed population is likely to be violated in 

 depletion experiments conducted with an otter trawl 

 in small geographic areas because of migration. 



Blue crabs in Chesapeake Bay are largely inactive 

 and bury themselves in the bottom sediment from 

 November through March (Van Engel, 1958); thus, 

 they are less likely to escape the dredge by swim- 

 ming. Orth and van Montfrans ( 1987) reported negli- 

 gible catches in bottom trawls during winter, further 

 supporting the premise that crabs are buried in the 

 substrate. Blue crabs captured in removal experi- 

 ments showed little signs of mobility when brought 

 aboard the vessel. We, therefore, believe that the 

 assumption of a closed population is fairly well met 

 during the short time span of each experiment in 

 winter. Also, fishing activity is at a minimum during 

 winter; only crabs in the Virginia mainstem of the 

 bay are harvested. 



We report on the catching efficiency of the samp- 

 ling dredge estimated from multiple removal experi- 

 ments in the blue crab survey. We demonstrate that 

 catchability estimates from a single or a few removal 

 experiments will not be reliable for the entire bay. 

 We show how the estimated catchability can be used 

 to calibrate the relative estimate of abundance from 

 the survey. 



Material and methods 



Removal experiments 



Maryland Department of Natural Resources (MDNR), 

 Chesapeake Biological Laboratory (CBL), and Virgi- 

 nia Institute of Marine Science (VIMS) conducted 88 

 removal experiments between November 1992 and 

 March 1995 using the standard 1.83-m wide Virginia 

 sampling dredge. The dredge was lined with either 

 12.7-mm hexagonal chicken wire or nylon mesh and 

 is assumed to have "knife edged" selectivity for crabs 

 with acarapace width (CW) of at least 15mm(Sulkin 

 and Miller, 1975). 



Depletion experiments generally were conducted 

 at locations that represent the variations in depth 

 and sediment type typical of Chesapeake Bay (Fig. 

 1), taking into account up-to-date survey data. It is 

 impractical to select sites for depletion experiments 

 randomly because blue crabs have a patchy distri- 

 bution, and the annual baywide dredge sui"vey gene- 

 rally has a large number of zero catches. Because of 

 cost considerations, removal experiments were con- 

 ducted each year at a random subset of survey sta- 

 tions with positive catches. 



Removal experiments were conducted within an 

 area of approximately 100 m by 5.5 m in Maryland 

 waters and 100 m by 9 m in Virginia waters. In both 

 cases experimental areas were marked with buoys. 

 Each removal from the experimental area (coverage) 

 consisted of three (CBL and MDNR) or five (VIMS) 

 parallel, nonoverlapping dredge tows conducted back 

 and forth (Fig. 2) at a standard towing speed of 3 

 knots. A maximum of 10 removals was completed for 

 each depletion experiment. The unit of effort was one 

 coverage (i.e. the combined 3 or 5 hauls required to 

 sweep the experimental area), and catch was recorded 

 as the total number of crabs caught per coverage. 



Estimating catchability 



Hirst (1994) formulated the following standard 

 assumptions for the removal method: 1 ) there is no 

 immigration to or emigration from the enclosed area 

 during the removal experiments; 2) each animal has 

 an equal probability of being caught; 3) each re- 

 moval is equally efficient (i.e. the probability of cap- 

 ture for each animal is constant from one removal to 

 the next). The first of these assumptions is reasona- 

 ble because crabs are largely inactive during winter, 

 and each depletion experiment is conducted over a 

 short time (2 to 4 hours). The latter two assumptions 

 may be less likely to be true because crabs generally 

 are clustered in distribution. In marine surveys the 

 sampling unit is typically a fixed volume, or a unit 



