In addition to the expressed deficiencies, 

 there is a disturbing hypothesis that all lob- 

 sters regardless of sex in any given age or 

 molt class may not shed in each year (Wilder, 

 1953 and Cooper, 1970). It follows then that 

 it would be meaningless to proceed further 

 with estimates from length frequencies of the 

 needed population parameters on age and 

 growth and mortalities. However, if we accept 

 the possibility of a fairly constant percentage 

 of an age or molt class shedding each year 

 over two or more years, then we have not 

 affected the estimates from the 14% groupings 

 that we need. In fact, Taylor (1948) stated 

 a similar premise in connection with converting 

 length groups to age groups. 



If the constant percentage premise were not 

 the case, I would expect the 14% increments 

 of carapace length compiled on a monthly 

 and yearly basis to be extremely erratic in 

 relation to each other. Of course, there are 

 other factors which might influence the fluc- 

 tuations in percentage from period to period, 

 such as sample size, effort, and year class 

 strength. Nevertheless, these fluctuations do 

 not mask certain characteristic patterns in the 

 size composition of the catch (Fig. 6). That is, 

 from year to year there is usually a gradual 

 increase in the percent frequency of the group- 

 ings from 81 through 92 mm for males, and 

 for females from August through December 

 in each year. Conversely, these same years 

 and groupings usually display a gradual de- 

 scendency from April through June. In this 

 case, I believe, the length frequencies ade- 

 quately portray the pattern of the size or molt 

 composition of the commercial population be- 

 fore and after shedding. 



In fact, the section on catch and effort sup- 

 ports the concept of shedding and resultant 

 recruitment influencing the length composition 

 of the catch. That is, as the monthly catch- 

 per-unit-of-effort values decline (April through 

 June) with increasing effort, the length fre- 

 quencies by 14% groupings from 81 through 

 92 mm also decline by month until shedding 

 and resultant recruitment occurs in July and 

 subsequent months; then the catch-per-unit- 

 of-effort values increase as usually does the 

 percentage of carapace lengths from 81 through 

 92 mm. 



We are also able to make general statements 

 about the fishery from these length frequencies. 

 For example, in the coastal waters of Maine 

 at least 60% (usually 80% or more) of the catch 

 by size and month occurs from 81- (3-3/16 

 inches) through 92-mm (3-5/8 inches) carapace 

 length. Even if we accept the possibility of 

 a segment of lobsters not shedding (at least 

 in the legal size range), the lobster industry 

 would be in immediate economic ruin because 

 it appears that most animals are caught soon 

 after recruitment from the sublegal to legal 

 size through shedding. 



I am compelled to note here that it is 

 almost inconceivable to work on a commercial, 

 long-lived species whereby over 80% of the 

 yearly catch is constrained within V^-inch inter- 

 val in carapace length. 



Analysis by probability paper. — Keeping 

 to the advisability of analyzing length frequen- 

 cies in different ways, we used probability 

 paper to pick out modes from the accumulative 

 percentages of carapace lengths of lobsters 

 that are captured by commercial and research 

 gear. The combination of the two types of 

 sampling allowed us to subject a wider range 

 of lobster lengths to the probability method 

 described by Harding (1949) and Cassie (1954). 



In this method, gear selectivity should be 

 considered for the two types of sampling be- 

 cause this factor alone may have an effect 

 on the location of the modes. Krouse (1971, 

 see footnote 3.) determined that wire traps 

 (1- x 2-inch and 1- x 1-inch mesh) have a 

 selective range down to at least 50-mm cara- 

 pace length and that lobsters appear to be 

 fully vulnerable between 68- and 70-mm cara- 

 pace length. As discussed previously, the com- 

 mercial gear possibly has a selective range 

 below the minimum legal size while the com- 

 mercial-sized lobsters appear to be fully vul- 

 nerable at 85-mm carapace length. This mode 

 might also coincide with an assumed age or 

 molt class. To suppoit this contention, we 

 found a similar mode for the catch from re- 

 search sampling gear (Krouse, 1971, see foot- 

 note 3.) It seems unlikely that this similar mode 

 in length frequencies from research and com- 

 mercial gear would occur by chance. 



24 



