Fishery Bulletin 94(1). 1996 



Gladius ageing analysis 



A total of 87 gladii were extracted from squid cap- 

 tured from the Navarin-Olyutorsky region during 

 June and July. The gladii were processed according 

 to the techniques elaborated by Bizikov ( 199 1 ). Each 

 gladius was dissected from the animal through the 

 dorsal section of the mantle, then washed carefully 

 with warm soapy water, labeled, and stored in 6% 

 formaldehyde. The periodic laminae were counted in 

 the cross sections of the inner shell layer of the 

 gladius by using the same microscope ( 100-200x) as 

 that used for statoliths. The sections were made at 

 the area of greatest thickness of the gladius inner 

 layer, near the ventral anterior margin of the conus. 

 These sections were placed on slides with a drop of 

 glycerine and covered with glass for viewing. In some 

 intact gladii, periodic laminae on the dorsal plate 

 (proostracum) were analyzed by using a zoom mi- 

 croscope under reflected light (32-56x magnification). 



Validation of the periodicity of increment 

 formation within ageing structures 



Two indirect methods were applied for validation of 

 the periodicity of statoliths and gladii increment for- 

 mation. The first involved comparison of increment 

 number in both gladius and statoliths taken from 

 the same specimen (Arkhipkin and Bizikov, 1991). 

 The second method included a comparison of the dif- 

 ference between the mean number of statolith incre- 

 ments from the modal groups of two successive 

 samples with the number of days elapsed between 

 these samples (Uozumi and Ohara, 1993; Uozumi and 

 Shiba, 1993). To analyze both relationships, a predic- 

 tive regression model was applied (Ricker, 1973). The 

 estimated values of slopes (SL) of both regressions were 

 tested with the null hypothesis (SL=1) at the 0.05% 

 level for SL by using the statistic t - (SL-D/SD, where 

 SD is a standard deviation of SL (Wesolowsky, 1976). 



Age and growth parameters 



Unfortunately, we had no opportunity to validate 

 directly the periodicity of growth increments within 

 statoliths and gladii. However, after examination of 

 increment number and microstructure within sta- 

 toliths taken from the same specimen and after com- 

 parison of these with daily growth increments in the 

 statoliths of validated species (Loligo opalescens, 

 Hixon and Villoch, 1984; Illex illecebrosus, Dawe et 

 al., 1985; Alloteuthis subulata, Lipinski, 1986; 

 Sepioteuthis lessoniana, Jackson, 1990; Todarodes 

 pacificus, Nakamura and Sakurai, 1990), we as- 

 sumed that growth increments were deposited daily 



(see the corresponding section in Results section). 

 We considered the total number of increments be- 

 yond the "natal ring" (observed by Natsukari et al., 

 1993) to be the age of squid in days. Hatching dates 

 were backcalculated. Month classes of hatching (fur- 

 ther "month classes") were defined by the pooling of 

 squid into each month of hatching. 



Length-at-age data were analyzed separately for 

 both sexes and hatching months. Monthly age struc- 

 ture in a given region was defined by routine proce- 

 dures of age-length key construction. For more pre- 

 cise construction of age-length keys, we determined 

 the numbers of squid of each month class separately 

 for each sex and maturity period (Table 1A). With these 

 keys, month-class compositions in each month were 

 estimated from the monthly length-frequency compo- 

 sitions (Table IB). The similarity of age structures ob- 

 served in different months was estimated by applying 

 a coefficient of similarity (as %, 0% as absolute differ- 

 ence, and 100% as absolute resemblance) elaborated 

 by Shorygin (1952) during studies offish food spectra. 



The relatively short period of investigation in this 

 study did not permit the resolution of ontogenetic 

 growth curves for each hatching month of squid. 

 However, it was possible to construct a growth curve 

 for the species by pooling the individual age-at-length 

 data over various hatching months. Curves (logistic, 

 Gompertz, and von Bertalanffy) were fitted to the 

 age-at-length data by using the method of iterative 

 nonlinear least squares. As a result, the values of 

 ML were calculated by using the formula of the best 

 fitted curve. Unfortunately, pitching and rolling of 

 the vessel during frequently occurring storms pre- 

 vented us from weighing precisely almost half of the 

 specimens analyzed, especially small-size ones. For 

 investigation of growth in weight, we constructed at 

 first length-weight curves for both sexes, and then cal- 

 culated age-weight curves. Daily growth rates in length 

 or weight (DGR; mm or g per day) and instantaneous 

 growth rates (G) were calculated after Ricker ( 1958) as 



and 



DGR = (M 2 -M 1 )/T 



G = (lnM 2 -lnM,)/T, 



where M, and M 9 are mantle length ( mm) or weight (g) at 

 the beginning and end of a time interval (T=30 days). 



Results 



Distribution and abundance 



Berryteuthis magister occurred in most trawl catches 

 along the continental slope of the western Bering Sea 



