CAMPBELL and EAGLES: SIZE OF ROCK CRABS 



gonadal development. Modifications to Haefner's 

 classification include combining Haefner's first and 

 second stages into a single immature stage and mi- 

 croscope measurements of oocytes, and examination 

 for spermatophores. The diameters of three oocytes 

 per ovary to the nearest 0.01 mm under a compound 

 microscope were measured. Female C. irroratus were 

 considered to be physiologically mature when 

 ovaries were developed to greater than or equal to 

 stage 3 since all ovigerous females had stage-3 

 ovaries. Male C. irroratus were estimated to be 

 physiologically mature as judged by the size of the 

 gonads and the presence of spermatophores (num- 

 ber per 0.1 mm 2 determined with a micrometer and 

 microscope) in the vas deferentia (Table 1). About 

 10% of the vas deferentia of each gonad stage was 

 subsampled.and examined for the presence of sper- 

 matophores under 400X magnification. All males 

 with gonad stage >3 had well-developed sper- 

 matophores. Stage-2 gonads were considered imma- 

 ture because of the low numbers of small 

 developing spermatophores. 



The proportion of physiologically mature crabs in 

 each sex was calculated by dividing the number of 

 mature gonad stages by the total number of gonads 

 examined for each 5 mm CW class. The relationship 

 between C W (X) in millimeters and proportion ma- 

 ture ( Y) for both female and male C. irroratus was ap- 

 proximated by the logistic function: 



Y = 



a 



1 + e b+cX 



where a, the asymptote of the curve, and b andc, em- 

 pirical constants, were estimated by nonlinear least 

 squares approximation from the C W and proportion 

 mature data using Marquardt's algorithm (Conway et 

 al. 1970; Marquardt 1963). 



Indices of functional maturity using morphometric 

 criteria for male and female rock crabs were obtained 

 by the following two methods: The first was to find 

 50% sexual maturity from the intersection of a pair of 

 linear regressions that had the best fit (Somerton 

 1980a, b) to CW-chela height data; the second by 



dividing the abdominal width by CW for each in- 

 dividual crab and averaging the ratios for each 5 mm 

 CW group. 



The power curve of the linear form log 10 Y = a + b 

 log 10 X was used to approximate the relationship be- 

 tween CW(X) and total weight in grams (Y), and the 

 number of eggs per female (Y), using the least 

 squares method. Analysis of covariance was used to 

 compare the slopes and elevations of all the regres- 

 sion equations (Snedecor and Cochran 1967). 



RESULTS AND DISCUSSION 



There were no significant differences using analysis 

 of covariance (P>0.05) in size at maturity, CW- 

 weight, or CW-fecundity relationships between rock 

 crabs of the same sex from each sample area; 

 therefore, the data were combined to present each 

 relationship for one general Bay of Fundy-south- 

 westernNova Scotia area. Table 2 indicates that the 

 CW-weight (g) relationship had a high correlation 

 coefficient (r) for male, nonovigerous, and ovigerous 

 C. irroratus. Although there were no significant dif- 

 ferences between the slopes (P>0.05), there were 

 differences in elevations (P< 0.001) of the CW- 

 weight regressions over similar CW ranges, between 

 ovigerous females and males and between ovigerous 

 and nonovigerous females (Table 2). Differences in 

 elevation between male and nonovigerous females 

 were probably not biologically significant at P< 0.05 

 (Table 2). Krouse (1972) and Scarratt and Lowe 

 (1972) also did not find a difference between males 

 and nonovigerous females in the CW-weight rela- 

 tionship. We, as did Scarratt and Lowe (1972), found 

 that ovigerous females were heavier than non- 

 ovigerous females carrying 2.2-45.2 g (wet weight) of 

 eggs for a 41.1-100.2 mm CW range. 



The CW-proportion mature relationship estimated 

 from gonad development was described well by the 

 logistic curve (Fig. 3). The C W at which 50% of males 

 and females were mature was estimated at 61.7 and 

 48.6 mm, respectively. The smallest male found with 

 mature gonads (stage 3) was 34.2 mm CW and the 

 largest male with immature gonads (stage 2) was 95. 1 



TABLE 2.— Regression constants for the carapace width (Y) and weight (X in grams) 

 relationship (log 10 Y = a + b log 10 X) for male, nonovigerous, and ovigerous rock crabs, Cancer 

 irroratus. r = correlation coefficient. 



' Elevations followed by same symbol were significantly different: * at P<0.05; o and + at P-C0.001 , but there were 

 no significant differences [P>0.Q^>) between all three slopes (b) using analysis of covariance. 



359 



