The dotted line representing those clams removed by bullraking begins 

 slightly below U5 mm. instead of just under 60 mm. as in the dredge area 

 which reflects the difference in the size composition of the catch by the 

 two methods as previously shown in Figure ?• Appendices A and C give 

 tables used in preparing Figures 8 - 11. Appendix C presents statistical 

 analyses for estimating sampling reliability. 



It would be desirable to know whether it is better to remove only 

 large clams as dredging does^ or to remove both large and small clams as 

 raking does. The present experiment, however, does not provide an answer 

 to this question, nor was this an original objective. We know that a 

 spawning stock must be left, but the magnitude of this stock and its size 

 composition has not yet been established. Further information is needed 

 on the annual mortality from causes other than man before we can decide 

 if growth from little neck to medium size will increase the yield suffi- 

 ciently to offset mortality. Economic considerations such as the price 

 differential betvreen little necks and mediums would affect a decision on 

 the best method of harvesting hard clams, but these factors are beyond 

 the scope of the present investigations. 



Disappearance of Group "B" in Control and in Bullraked Areas 



Figures 12 and 13 show the size composition of the population in 

 autumn 19h9 and 1950 in the control area. A great change occurred in 

 this area even though we removed no clams . The group of clams from 30 to 

 56 mm. in Figure 12, which we will designate as Group "A", decreased 

 19.0^ by 1950 as determined from clamshell bucket samples. The larger 

 group from 57 to 75 mm. in 19U9, which will be known as Group "B", 

 decreased 70,55^ by 1950. The combined groups had a loss of 35.7^» 



The original presence of Group "B" is substantiated by sampling of 

 the test plot in May 19U9 with a dredge equipped with a liner in the bag 

 to retain small clams. At that time this group ranged from 52 to 70 mm. 

 and comprised 35.3:^ of the total as shown in Figure H. In the Novem- 

 ber 1, I9U95 survey (Fig. 12), Group "B" had grown to 57-75 mm. and 

 included 30.3?^ of the total. By November 8, 1950 (Fig. 13) it had grown 

 to 6U-79 mm. but contributed only 13.145S of the population. Duplicate 

 sampling in 1950 substantiated the disappearance of Group "B" as shown in 

 Appendix G. 



Statistical analyses of the differences in mean number of clams per 

 sample between 19U9 and 1950 showed the probability of this difference 

 occurring by chance is less than one time in 100. (Appendix C) This 

 means there was a real difference in the population of the control area 

 in the two years 5 that this difference was not due to sampling error. 



During this same period Group "B" had largely disappeared from the 

 bullraked area also. Catch measurements of clams bullraked from each of 

 the four quarters in 19U9 showed the presence of this larger group, Bull- 

 raking was completed September 30, 19U9. The clamshell bucket census of 



21 



