350 



Fishery Bulletin 94(2). 1996 



by first air drying the egg masses and, after separat- 

 ing the eggs from the pleopods and setae, weighing 

 them to the nearest 0.1 mg. Two subsamples of about 

 200 eggs from each dried egg mass were randomly 

 selected, weighed, and counted. Fecundity was then 

 estimated by dividing the total weight of an egg mass 

 by the average of the two estimates of individual egg 

 weight. 



Male functional maturity (Conan and Comeau, 

 1986) was determined from chela allometry. The pro- 

 cedure used is based on the observation that, for 

 Chionoecetes, when chela measurements are plotted 

 against carapace measurements on log-log axes, the 

 data conform to two, nearly parallel, straight lines, 

 one representing immature individuals and the other 

 representing individuals that have undergone a pu- 

 berty molt and have acquired the secondary sexual 

 characteristic of enlarged chelae (Hartnoll, 1978; 

 Somerton, 1980; Conan and Comeau, 1986). We de- 

 termined with which line an individual observation 

 was associated by using the computer algorithm de- 

 scribed in Somerton ( 1980) to reassign observations 

 iteratively to the two lines until the residual sum of 

 squares was minimized. The carapace width at 50% 

 maturity W 50 ) for both males and females was esti- 

 mated by fitting a logistic regression model to the 

 maturity and carapace width data by using general- 

 ized linear modeling (Chambers and 

 Hastie, 1992). W.,, and the variance of 



over the latitudinal range of the survey area (Fig. 2), 

 but fewer hauls were made at greater depths. Over 

 the survey area, both species displayed patchy dis- 

 tributions; much of the catch came from relatively 

 few sampling sites (Fig. 2). Chionoecetes tanneri oc- 

 curred at 47 sites and had a mean CPUE (catch in 

 numbers per kilometer of towed distance) at these 

 sites of 1.9; Chionoecetes angulatus occurred at 50 

 sites and had a mean CPUE of 2.6. The difference in 

 CPUE was not significant U-test, P=0.418). 



Size distribution 



Chionoecetes tanneri males ranged 42-170 mm in 

 carapace width (CW); females ranged 38-126 mm 

 CW (Fig. 3). The maximum size of females was simi- 

 lar to maximum sizes reported off Oregon ( 126 mm, 

 Pereyra, 1966; 122 mm, Tester and Carey, 1986), 

 whereas the maximum size of males fell between the 

 maximum sizes reported for Oregon (181 mm, 

 Pereyra, 1966; 162 mm, Tester and Carey, 1986). 

 Chionoecetes angulatus males ranged 25-160 mm 

 CW; females ranged 35-104 mm CW. The maximum 

 sizes of both sexes were slightly larger than the re- 

 ported maximum sizes of C. japonieus in the Sea of 

 Japan (males 150 mm, females 90 mm, Watanabe 

 and Suzuuchi, 1983). 



50 



W 50 were then estimated from the pa- 

 rameters of the fitted model with formu- 

 las provided in Somerton ( 1980). In addi- 

 tion to W 50 , we also calculated the mean 

 carapace width of mature individuals 

 ( W ) as an alternative measure of the size 

 at maturity for comparative purposes. 



Weight and carapace width relation- 

 ships for males of each species were es- 

 timated by nonlinear regression (Cham- 

 bers and Hastie, 1992) to fit a power 

 function to weight and carapace width 

 data. A between-species difference in 

 these relationships was tested by mak- 

 ing the relationships linear with a log 

 transformation of the data, and by us- 

 ing analysis of covariance to test whether 

 species-specific relationships fit the 

 data significantly better than a single, 

 combined relationship. 



Results and discussion 



The 205 trawl hauls that were success- 

 fully completed were distributed evenly 



\ All hauls 



|} C. tanneri 



(2 C angulatus 











200 400 600 800 



200 400 600 800 



Depth (m) 





200 400 600 800 



Figure 2 



Distribution of all successful trawl hauls (Ai, positive Chionoecetes tanneri 

 hauls iBi. and positive C. angulatus hauls (Cl. with respect to latitude 

 and depth in meters. For both species, the size of the circle is proportional 

 to the total catch in numbers per kilometers of towed distance. 



