was reported at the Intake station following dredg- 

 ing in 1985 (NUSCO 1987a). Dredging in the 

 vicinity of the intake structures removed existing 

 habitat (shelters), and catches at Intake during 

 1985 were the lowest reported for that station 

 since wire pots were first used. After the dredged 

 area stabilized, lobsters returned to the area and 

 catches at Intake increased in 1986 and 1987. 

 Pottle and Elner (1982) reported that smothering 

 of habitats occupied by juvenile lobsters may re- 

 sult in increased spatial competition among lob- 

 sters for shelter in remaining gravel habitats. 



In addition to lobsters, trap catches contained 

 several species of vertebrates and invertebrates. 

 In 1984, we began analyzing the effects of these 

 species on lobster catch using covariance analysis 

 (NUSCO 1985). The species caught in pots at 

 each station were Initially used as covariates to 

 identify species that significantly (p < 0.05) influ- 

 enced lobster catch (Table 3). During 1986 and 

 1987, the influence of catches of competing species 

 on lobster CPUE was the same as in previous 

 years. Five species were identified as significantly 

 influencing lobster CPUE (spider crabs, hermit 

 crabs, whelks, summer and winter flounder). The 

 whelk or conch {Busycon spp.) was the only spe- 

 cies that had a positive influence on lobster catch. 

 Soaktime was another parameter used in the 

 model as a covariate to examine the significance 

 of varying set-over-days on lobster catch at each 

 station. Soaktime had a significant influence on 

 lobster catch only at Intake; during 1986 and 

 1987, however, soaktime significantly influenced 

 lobster catch at all stations during 1984. Our 

 soaktimes were not very different; the majority of 

 pots were hauled after a two or three day set. 

 Commercial lobstermen, however, vary the num- 

 ber of days between pot hauls to maximize their 

 catch; therefore, catch statistics for the commercial 

 lobster fishery are often based on catch per trap 

 haul set-over-day to account for the variability in 

 soaktime. The average soaktime for the (Connect- 

 icut fishery has ranged between 3.2 and 4.5 days 

 since 1978, whereas our soaktime averaged be- 

 tween 2.4 and 2.6 days over the same period (CT 

 DEP Marine Fishery Statistics; NUSCO 1987a). 

 The mean CPUE adjusted for the significant 



covariates (least square means) are presented with 

 its unadjusted arithmetic mean in Table 2 for 

 each station. Given the similarity between the 

 arithmetic and adjusted mean, the incidental catch 

 of competing species and the variability of our 

 soaktimes did not influence the reliability of our 

 CPUE. 



Population Characteristics 

 Size Frequencies 



Annual size frequency distributions for male 

 and female lobsters caught in wire pots from 1979 

 to 1987 are shown in Figure 3. The mean 

 carapace lengths (CL) of lobsters caught during 

 1986 (70.1mm) and 1987 (70.2mm) were smaller 

 than previous years' range 1978-85 (70. 7-7 1.8mm) 

 (Table 4). When 3 stations were compared, the 

 mean CL's of lobsters caught at .lordan Cove and 

 Intake were within the range of values reported 

 since 1978, however, the mean CL at Twotree 

 during 1986-87 was smaller than previous years 

 (Table 5). From 1978 to 1984, catches at Twotree 

 contained larger lobsters and greater proportions 

 of legal-size lobsters than nearshore catches; since 

 1985, the mean CL of lobsters and percentage of 

 legals caught at Twotree have been similar to the 

 nearshore stations. The mean size and proportion 

 of legal-size lobsters caught in our studies since 

 1978 were lower than that reported by other stud- 

 ies in LIS (Smith 1977; Briggs and Mushacke 

 1979; Marcello et al. 1979). 



Sex Ratios 



Sex ratios of males to females were 1.0:0.87 and 

 1.0:0.88 during 1986 and 1987, respectively, which 

 were within the range of values reported in pre- 

 vious years (Table 6). The Twotree catch had a 

 consistently higher proportion of females than 

 Intake and Jordan Cove catches, which contained 

 more males. This trend in sex ratios has been 

 consistent at the three stations since the study 

 began. Smith (1977) found male to female ratios 

 of commercial catches ranging from 1.0:1.06 to 

 1.0:1.81 in four different areas of LIS, which is 

 similar to ratios at Twotree 1.5 km offshore. Sex 



126 



