The length frequencies by sex of the com- 

 mercial catch are similar; therefore, we com- 

 bined these data for the probability analysis. 

 To further examine the assumption regarding 

 the similarity of the size composition between 

 the sexes, we simply plotted the accumulative 

 percent frequencies by sex on probability paper 

 by month and then year. The inflexion points 

 are approximately the same, indicating that 

 the probability method would yield almost iden- 

 tical modes. 



At first this situation seems to be in conflict 

 with the expectation that mature females extrude 

 their eggs in one year and usually carry them 

 externally into the next year before these eggs 

 hatch and the female possibly molts. The elapsed 

 time for nonshedding of mature females could 

 be 18 or more months. Therefore, with a cer- 

 tain percentage of males shedding each year 

 and a regulation protecting "v" notched or 

 berried females, there should be a difference 

 in the size composition between males and 

 females. The section on berried female mea- 

 surements helps to explain this apparent anom- 

 aly, in that those length-frequency data lead 

 me to believe that the majority of native females 

 are caught before they extrude eggs. This 

 situation could account for the similarity in 

 the length frequencies by sex in the commercial 

 catch. 



The probability method on the length fre- 

 quencies of the commercial catch by year re- 

 vealed similar curves for 1967, 1968, 1969, 

 and 1970 (Fig. 7). With this similarity, we should 

 expect the resultant modes in millimeters (cara- 

 pace length) to be approximately the same 

 from year to year (Table 6). 



As mentioned earlier, we calculated an 

 average of 8% per molt from laboratory animals. 

 The consecutive probability modes from the 

 commercial catch do compare favorably with 

 this 8% increment. For example, in 1967 the 

 percent increments between modes are: 7.1%, 

 6.6% , 8.2% , and 5.7% while in 1970 the percent 

 increments are: 8.3% , 4.4% , 11.6% , and 3.8% . 



I am reluctant to postulate that these con- 

 secutive increments actually portray the growth 

 pattern between age groups of lobsters in 

 the commercial catch. Still, these consecutive 

 modes may be the result of some situation 

 that I have overlooked. Confounding the prob- 

 lem even more, these modes give logical esti- 



mates of mortality and of parameters in the 

 von Bertalanffy Growth Equation. 



Comparison of 14% increments with proba- 

 bility modes. — The consecutive modes from 

 the probability analysis do not fall within the 

 successive ranges of 14% groupings in length. 

 However, we reasoned that it is unlikely for 

 the initial sizes of the range in length about 

 the 84- to 85-mm probability mode (assumed 

 age or molt class) to begin at the legal minimum 

 size of 81-mm carapace length. In fact, three 

 standard deviations about the 84- to 85-mm 

 probability mode extends the size well below 

 81 mm. Coupled with this, there could be a 

 range of sizes of a sublegal assumed age or 

 molt class extending into the protected size 

 range of the probability mode at 85 mm. If 

 this were true, then we would have a conglom- 

 erate of assumed age and molt classes in 

 subsequent years in the commercial fishery. 



Undaunted by this seemingly incongruous 

 situation, we attempted to follow the 85-mm 

 mode and its protected and unprotected size 

 range by approximate 14% increments from 

 1967 through 1969. This increase should be 

 the result of shedding. Therefore, the 85-mm 

 mode in 1967 might result in a mode at 97 mm 

 in 1968 while the protected size ranges of this 

 or another assumed molt class might move from 

 the sublegal sizes in 1967 to produce a mode 

 at 91 mm in 1968. The 97-mm mode in 1968 

 might move to 113 mm in 1969, while the 

 mode at 91 mm in 1968 might move to 102 mm 

 in 1969. 



If this were the actual situation, then the 

 modes from the probability analysis do agree 

 with the 14% groupings (listed in parentheses) 

 in the following manner: 85-mm mode (81-92 

 mm), 97-mm mode (93-106 mm), and 111-mm 

 mode (107-122 mm). The additional modes near 

 91 and 105 mm could be the result of the mini- 

 mum size regulation. 



Viewing the relationship between the two 

 techniques in another way, we hypothesized 

 that the 14% grouping from 81 through 92 mm 

 includes two probability modes at 85 and 91 

 mm; the grouping from 93 through 106 mm 

 includes two probability modes at 97 and 

 105 mm; the grouping, with a small sample 

 size, from 107 through 122 mm includes a 

 probability mode near 111 mm. Then this com- 



25 



