V0lstad et al Estimating dredge catching efficiency for Callinectes sapidus 



417 



for blue crab with CW >15 mm was T^= 

 5.86xl0'^(/t=:0.09). and 2.58xl0«(^=0.10) for 

 crabs with CW>60 mm. Using q-0.15 with 

 standard error 0.02 (from model 2), we esti- 

 mated absolute abundance for crabs with 

 CW >15 mm to be f2 = 6.26xl0« (>t=0.15), 

 and 2.75x108 (/fe=0.16) for crabs with CW 

 >60 mm. To check plausibility of these esti- 

 mates, we compared the absolute abun- 

 dance estimates of the 1+ age group (CW 60 

 mm) with estimated total landings for 1995. 

 Blue crabs of age 1+ reach harvestable size 

 (127 mm) the following fishing season. The 

 reported total landings in the commercial 

 fishery were 17,820 metric tons (t). Using an 

 average weight per crab of 150 g (Knotts^), 

 we estimated the total number of crabs 

 caught to be 1.19x10*^ and the exploita- 

 tion coefficient (as a ratio of catch in num- 

 bers and number of age 1+ crabs) to be 

 about 45%. 



Exploitation rates can be calculated for 

 males and females separately in similar 

 fashion by using mean CPUE and landings 

 by sex. These rates may be very valuable 

 infoiTnation because there is an evident 

 disproportion in crab landings by sex; on 

 average more males are landed per year (at 

 least by weight ) than females ( Rugolo et al. , 

 1998a). For example, 9320 t of males and 

 7230 t of females were landed in Maryland 

 in 1995. If the average weights of males 

 and females in the catch were similar, the 

 exploitation rate for males would be higher 

 than that for females. However, uncorrec- 

 ted for catchability, the density of age 1+ 

 males (2.31 per 1000 m^) observed in 1995 

 was higher than that of age 1+ females 

 (1.44 per 1000 ni'^), suggesting that fema- 

 les are being exploited at a higher rate. To 

 obtain precise estimates of sex-specific exploitation 

 rates, data on mean weight of crabs by sex in the har- 

 vest are required but are not available. 



Discussion 



Our method for estimating overall dredge catching 

 efficiency provides consistent estimates over time ( Fig. 

 3). The Leslie model produces the most precise esti- 

 mate for catching efficiency, but the estimate could 

 be slightly biased upwards if measurement error in 



0.25 



•: 02 



15 



1 



0.05 



1993 



1994 1995 



year 



n model 1 □ model 2 



all 



Figure 3 



Jackknife estimates of average catchability (?) by year for model 1 and 

 model 2. The error-bars represent the half width of the confidence intervals. 



^Knotts.K. S. 1989. Prehmmary stock assessment of the Ches- 

 apeake Bav blue crab population. M.S. thesis, Univ. Marvland, 

 College Park, MD, 206 p. 



effort occurs. In practice, given that the standard unit 

 of effort in the depletion experiments is one complete 

 coverage of the closed area, the three to five tows for- 

 ming a coverage may partially overlap or extend out- 

 side the area because of navigational errors or effects 

 of bottom cun-ents. This measurement error in catch 

 or effort inflates the average catchability coefficient 

 estimates from the Leslie model (Gould et al., 1997). 

 For model 2, the estimate of q for individual experi- 

 ments depends on the regression slope and its vari- 

 ance and will decrease with increasing variance. The 

 method for estimating q for individual experiments 

 assumes that the distribution of the estimated slope 

 tends toward normality; therefore estimates of based 

 q on model 2 could be biased at low sample sizes, but 



