FISHERY BULLETIN: VOL. 77, NO, 4 



We observed a general relationship between the 

 mean size of lobster caught and fishing location. 

 Comparisons of the mean size of lobster in sample 

 catches (pooled by trap type) for six fishermen, for 

 which adequate data were available, revealed a 

 segregation by fishing location (Table 4). In gen- 

 eral, lobster taken in Narragansett Bay and near- 

 shore Rhode Island Sound samples were sig- 

 nificantly smaller ( <^ = 0.05) than those taken in 

 offshore Rhode Island Sound when compared 

 using Duncan's multiple range procedure (Steel 

 and Torrie 1960), although one offshore sample 

 did not conform to this pattern. We attributed the 

 smaller mean size in Narragansett Bay and near- 

 shore Rhode Island Sound samples to intense 

 fishing pressure in these easily accessible areas. 

 Krouse (1973) noted a similar correspondence be- 

 tween fishing intensity and size composition of the 

 catch. Areas within Narragansett and Rhode Is- 

 land Sound with the smallest mean size of lobster 

 also had the lowest CPUE (Table 4). 



Characteristics of the habitat may also 

 influence the size composition of the catch. Several 

 authors have observed a correlation between the 

 size of lobster and the size of available shelter sites 

 (Scarratt 1968; Cobb 1971; Stewart 1972). Larger 

 lobsters were found in areas with greater shelter 

 size (Scarratt 1968; Cobb 1971) or in mud areas 

 with a high clay fraction capable of supporting 

 larger burrows (Stewart 1972). Inshore rocky 

 habitats are characterized by ledge and mixed 

 rocky debris which offer smaller shelter sites than 

 offshore mud and rock substrates. 



Table 4. — Results of Duncsin's multiple range procedure com- 

 paring mean carapace length{rankordered)of American lobster 

 from offshore Rhode Island Sound (R.I.S.), nearshore Rhode Is- 

 land Sound (R.I.S.N) and Narragansett Bay (N.B.). Means with 

 the same letter code are not significantly different ( <i = 0.05). 



Sex Ratios 



Comparisons of sex ratios in vented and control 

 traps revealed interesting differences. We noted a 

 female:male ratio of 1.68: 1 in non vented traps and 

 2.15:1 in vented gear. Contingency table analyses 

 indicated that the sex composition of the catch 

 differed significantly in vented and control traps 



(X,^ = 7.70;P<0.01). These data suggest differen- 

 tial escapement by males and females. To further 

 assess this possibility, we investigated the rela- 

 tionship between carapace length and carapace 

 width for 437 male and 603 female lobster. 

 Analyses of covariance (Steel and Torrie 1960) 

 indicated that the regression coefficients were 

 significantly different (Fi^^^g = 6.74; oc = Q.Ol). 

 The least squares regression equations were 



CW„, = -0.8901 + 0.6186 CL„, (r = 0.869) for 



males and 

 CWf = -4.3932 + 0.6755 CLf (r = 0.886) for 



females. 



In passing through a rectangular vent, the criti- 

 cal body dimension is the carapace width (the 

 minimum body measure). The relatively broader 

 carapace width for females of a given carapace 

 length may result in the retention of proportion- 

 ately more females, accounting for the observed 

 discrepancy in sex ratios in the experimental gear. 

 It should be noted that Krouse and Thomas ( 1975) 

 found no significant differences in the carapace 

 width-length relationship for 114 female and 103 

 male lobster. 



Vent Size, Orientation, 

 and Lath Spacing 



We examined the effect of vent orientation 

 (horizontal vs. vertical) and lath spacing on es- 

 capement. The effectiveness of larger vents (44.5 

 mm X 152 mm) in retaining legal lobster was also 

 tested. Vent orientation may affect either the 

 probability of a lobster locating the vent or the 

 time required to find the vent, a factor of impor- 

 tance with short immersion times. There may also 

 be differences in size selectivity associated with 

 vent orientation. Analysis of preliminary size 

 composition data indicated that 42 mm vents may 

 in fact be too small for a minimum legal size of 78 

 mm. Accordingly, we tested 44.5 mm x 152 mm 

 vents in an attempt to determine if legal-sized 

 lobster could escape with the use of this larger 

 vent size. We also evaluated the effectiveness of 

 opening lath spacing to 42 mm and 44.5 mm in 

 comparisons with vents of equivalent size. 



A total of 4,487 lobster were obtained in 2,222 

 trap hauls of the experimental gear. As might be 

 expected, traps with 44.5 mm openings (vented 

 and lath spaced traps) retained markedly fewer 

 sublegal lobster than either traps with 42 mm 



930 



