202 



Fishery Bulletin 100(2) 



their daytime equivalent. The size groups (<80 mm ML, 

 >80 mm ML) were chosen to allow comparisons with re- 

 sults from previous studies (e.g. Lange, 1980: 1981; Lange 

 and Sissenwine, 1983: Brodziak and Hendrickson. 1999: 

 NEFC': Lange and Sissenwine-: Lange'^). Alternative anal- 

 yses were performed for different size groups (<50 mm 

 ML, >50 mm ML and <100 mm ML, >100 mm ML) to 

 assess the sensitivity of the results to the choice of size 

 groups. The combined effects of geographic region (NE and 

 MAE), depth zone, and year on sui-vey catches were tested 

 by using generalized linear models (GLM) to derive main 

 effects and coefficients for each sui-v^ey. Paii^wise compari- 

 sons were tested by using a Mest with Bonferroni adjust- 

 ments (Sokal and Rohlf, 1995) to compare specific regions, 

 seasons, and depth zones. All tests were analyzed at the 

 5'y( significance level. Differences between seasons and re- 

 gions were tested between autumn and spring sui'veys for 

 the years 1968 to 1997, between autumn and winter for 

 1992 to 1997, and between spring and winter surveys from 

 1992 to 1998. Proportion of catches <50 mm ML were ana- 

 lyzed to evaluate the relative distribution of juvenile L. 

 pealeii. 



Biological analysis 



Subsamples of 50-100 individuals were obtained from five 

 different survey time series: NEFSC autumn (September- 

 October 1997), winter (February 1998), and spring (March 

 1998), inshore Massachusetts (Howe^) (October 1997), and 

 Connecticut (Johnson") (Long Island Sound, May 1998). 

 The samples were analyzed from each of five depth zones 

 (1-26 m, 27-55 m, 56-100 m, 111-185 m, 186-366 m), 

 within each of three geographic regions (Gulf of Maine 

 [GOM): Georges Bank-Southern New England, north of 

 Hudson Canyon |SNE|: and Mid-Atlantic Bight, see above 

 [MABl). A fourth region, south of Cape Hatteras (SOH) 

 was added later. Each sample comprised a nonrandom 

 selection of lengths to represent the size range present in 

 a tow. In total, 2156 individuals were subsampled from 53 

 sun'ey tows. Sexes were determined and specimens were 

 measured to enable the morphometric maturity analyses 

 of Macy (1982): each individual squid was also weighed on 

 a top-loading balance to 0.1 g. The morphometric method 

 uses a suite of length measurements for female and male 

 squid to determine maturity stage (measured on a scale of 

 1 to 4, where 1 is immature and 4 is fully mature). Oppor- 

 tunistic commercial samples from early winter (December 

 1998 and January 1999) were also analyzed (118 individu- 

 als) to bridge the temporal gap in sui-vey coverage. 



Data on dorsal mantle length (ML, mm) and total body 

 mass (BM, g) for each maturity stage were used to esti- 

 mate proportions for each maturity stage across the length 



^ Howe, A. B. 1989. State of Massachusetts inshore bottom 

 trawl survey. Atlantic States Marine Fisheries Conimi-~sion 

 (ASMFC) Spec. Rep. 17:33-38. [Available from ASMFC, 1444 

 Eye Street, N.W., sixth floor, Washington. DC 2000.5.] 



" Johnson, M. 1994. State of Connecticut marine finfish trawl 

 survey. Atlantic States Marine Fisheries Commission (ASMFC ) 

 Spec Rep. 35:24-26. [Available from ASMFC, 1444 Eve Street, 

 N.W., sixth floor, Washington, DC 20005.1 



and weight range. These proportions were used to deter- 

 mine the sizes at which squid of both sexes changed from 

 one maturity stage to the next. 



Maturity-at-length data were weighted by diurnally ad- 

 justed catch-at-length data for each depth zone and region to 

 provide population-weighted maturity patterns, assuming 

 that sui"vey length distributions accurately represent rel- 

 ative proportions of population components. Catch-weight- 

 ed data were analyzed to derive 1 ) the patterns of matu- 

 rity for each sex at different times of the year: 2 ) estimates 

 of proportions of each maturity stage sampled by survey: 

 3 ) mean length for each maturity stage of each sui-vey: 4 ) 

 mean length for each region of each survey: and 5) mean 

 length for each depth stratum of each survey. A small pro- 

 portion (8.4^^ ) of survey catches <50 mm ML were not sub- 

 sampled: these were assigned to the juvenile stage. Catches 

 at larger sizes, which were not subsampled (6.8% of sui-vey 

 catches), were removed from the analysis because sex or 

 maturity stage could not be assigned with any degree of cer- 

 tainty Individual squid, or size classes of squid, were not 

 weighed during NEFSC surveys: therefore the maturity da- 

 ta from the biological analysis could only be catch-weighted 

 by length because length was measured on a random sub- 

 sample of squid caught at each station in NEFSC surveys. 



Results 



Survey analysis 



Patterns of diurnal distribution were different among sea- 

 sons surveyed (Table 1). In winter surveys, from 1992 to 

 1998, prerecruit (i.e. <80 mm ML) catch was lower at night 

 and during dawn and dusk than during daylight hours 

 (65'/J and 81% of daytime catch, respectively). However, for 

 recruits (i.e. >80 mm ML), catch was higher both at night 

 and at dawn and dusk than during the day ( 131% and 115% 

 of daytime catch respectively) in winter sui-veys. In autumn 

 and spring sui'veys, from 1968 to 1998. both prerecruits and 

 recruits showed a lower catch at night and during dawn 

 and dusk than by day: recruits showed a lesser diurnal 

 variation than prerecruits. Results from analyses with dif- 

 ferent size groups (<50 mm ML, >50 mm ML and <100 mm 

 ML, >100 mm ML) were very similar, suggesting that the 

 interaction of size and time of day is gradual. 



Catch rates varied significantly by season (Table 2). 

 During winter and spring, survey catches were greater in 

 the MAB by a factor of approximately four (Fig. 2). How- 

 ever, there was no significant difference in sui'vey catches 

 between geographic regions in autumn (Fig. 2). Pairwise 

 comparisons showed that mean number-per-tow was sig- 

 nificantly greater in autumn than in spring within both 

 geographic regions and was greater in autumn than in 

 winter in the NE. There were no significant differences be- 

 tween autumn and winter means in the MAB nor between 

 spring and winter means in either the MAB or the NE. 



Catch by depth, pooled over the MAB and NE, varied by 

 season (Table 3). Pairwise comparisons of each depth for 

 each season showed that winter and spring survey catch- 

 es were lowest in the shallowest stratum (27-55 mi, in- 



