line model is not significantly better than the fit of a 

 single line, either more data or a broader size range of 

 data is needed or some other procedure for estimat- 

 ing the size of maturity should be considered. 



The standard deviation of X* was estimated using 

 Monte Carlo simulation (Hammersley and Hands- 

 comb 1964), which, in this particular application, 

 consisted of 1) generating sets of synthetic mor- 

 phometry data, 2) estimating X* for each data set by 

 fitting the two line model, and 3) calculating the stan- 

 dard deviation between the estimates of X*. Each 

 synthetic data set was constructed by generating a 

 new chela measurement for each carapace measure- 

 ment in the original sample. The new chela measure- 

 ment was computed as 



Y = E{Y) + Z • SD 



where Y is the logarithm of chela height, E(Y) is the 

 expected value of log chela height given carapace 

 length and the parameters of the appropriate phase 

 line, Z is a randomly generated standard normal de- 

 viate, and SD is the standard deviation about the ap- 

 propriate phase line. For carapace widths <X*, the 

 parameters and SD for the juvenile phase line were 

 used; for carapace widths >X*, the parameters and 

 SD for the adult phase line were used. Thirty samples 

 of morphometric data were generated for each pop- 

 ulation. X* was estimated for each sample and the 



standard deviation among the 30 independent es- 

 timates of X* was calculated. The estimates of X*, 

 standard deviations, and 95% confidence intervals 

 are shown in Table 2. 



Samples from the populations of St. Matthew Island 

 and Pribilof Islands contained some chela measure- 

 ments that were unusually small compared with 

 other measurements from crabs with similar car- 

 apace lengths (Fig. 4). These atypical measurements 

 were probably obtained from crabs in the process of 

 regenerating lost chelipeds. For crabs which are bi- 

 laterally symmetric, the inadvertent measurement of 

 partially regenerated chelae should not be a problem, 

 because the sizes of left and right chelae can be com- 

 pared in the field, and, if different, the measurement 

 can be rejected. For crabs, such as blue king crab, 

 which are not bilaterally symmetric, partially regen- 

 erated chelae are harder to detect and thus are likely 

 to be included in the sample. For red king crab, the 



Table 2. — Male size at maturity, standard deviation, and 95' i con- 

 fidence intervals for each of the four blue king crab populations 

 studied. 



100 r 



- 80 



z: 



5 60 



»— » 



LJ 



cr 



_i 



X 



40 



20 



N=834 



40 80 120 160 



CARAPACE LENGTH (MM) 



FIGURE 4.— Male chela height and carapace length 

 measurements of the blue king crab from the Pribilof 

 Islands before date deletion. Measurements of chela 

 height which appear unusually small relative to the 

 carapace length probably were obtained from crabs in 

 the process of regenerating lost chela. 



200 



625 



