Quihonez-Velazquez: Age validation and growth of Melanogrammus aeglefinus and Po/lachius virens 309 



pollock larval otoliths from the validation experiment 

 and from wild fish <36 mm SL were mounted on glass 

 over slides with Permount (Fisher Scientific Labo- 

 ratory, Fair Lawn, NJ). Eighteen pairs of haddock 

 otoliths from larvae measuring 50-83 mm SL were 

 mounted with synthetic resin on metal stubs for SEM 

 observation. 



The mounted otoliths (sagittae) used in the vali- 

 dation experiment were read with an image analyzer 

 system by using Optimas software (Subtechnique, 

 Inc., Alexandria, VA). Otoliths from field-collected 

 larvae and juveniles were read at the Southwest Fish- 

 eries Center (La Jolla, CA) by using the OTO pro- 

 gram (Andersen and Mokness, 1988). The image 

 analyzer system consisted of a video camera attached 

 to a compound microscope, monitor, digitizer, and 

 microcomputer. A detailed description of the OTO 

 program, otolith analyzing system, and methods used 

 are given by Andersen and Moksness ( 1988 ). Otoliths 

 examined with SEM were ground on the sagittal 

 plane to the nucleus with fine grit paper (between 

 30 |am and 0.3 |am), and then the polished surface 

 was etched for 6 min in 0.2M EDTA (pH 7.6). 



In addition to the number of increments, the fol- 

 lowing otolith measurements were taken and used 

 in comparisons of each species: maximum otolith di- 

 ameter (OD); maximum otolith radius from the cen- 

 ter of the nucleus to the outer edge of the otolith (OR); 

 diameter of the otolith nucleus (ON); and diameter 

 of the yolksac resorption check (YSC) as defined by 

 Radtke and Waiwood (1980) and Bolz and Lough 

 (1983). Growth increments were counted twice by the 

 same reader at an interval of 3 mo between read- 

 ings. The two counts usually differed <69'f, and the 

 average was used as an estimate of age. 



Statistical procedures 



For both species, of all otolith measurements, the 

 maximum otolith diameter was most strongly corre- 

 lated with length. I compared the relation of otolith 

 growth to larval growth between month and year for 

 both species to determine whether otolith growth could 

 be used to predict lai-val growth. Thus, linear regres- 

 sions of LnSL on LnOD were compared by using analy- 

 sis of covariance ( ANCOVA) with OD as the covariate. 

 A Laird-Gompertz growth model curve was fitted 

 to the length-at-age data for each species in each year. 

 This model has been shown to provide an adequate 

 fit for length-at-age data on age 0+ fish of many dif- 

 ferent species (Bolz and Lough, 1988; Lough et al., 

 1982; Watanabe et al., 1988; Simard et al., 1992). 

 Zweifel and Lasker ( 1976) presented a detailed dis- 

 ' cussion of the Laird-Gompertz function. The equa- 

 tion for the model is 



Lt 



L^e 



kd-e"") 



where L„ 



L. - 



"0 = 



length at ^=0; 



a dimensionless parameter, such that 



ka=AQ is the specific growth rate at /=0 



(A,=Aoe^"; 



length at any age t; 



the specific rate of growth when t-t^^; 



and 



the time when the growth rate starts to 



decrease, that is, the inflection point of 



the curve (Ricker, 1979). 



The parameters were derived by nonlinear least 

 squares tests by using the SYSTAT nonlinear pro- 

 gram (Wilkinson, 1990), and ANCOVA was used to 

 test whether differences in growth rates in different 

 years of each species were significant. 



Results 



Validation of daily increment deposition 



In both species, otoliths (sagittae and lapilli) were 

 present at hatching. Some haddock otoliths ( Fig. 2A) 

 showed one to three irregular increments between 

 the nucleus and the check at hatching, as observed 

 by Bolz and Lough (1983). The mean (±SD) dia- 

 meters of the nucleus and of the check at hatching of 

 reared larvae were 10.8 ±2.5 and 21.4 ±3.3 |am, re- 

 spectively, in haddock and 12.0 ±1.9 and 20.7 ±2.2 

 Hm, respectively, in pollock. The values for haddock 

 were consistent with earlier reports by Bolz and 

 Lough (1983) and Campana (1989). In both species, 

 increments appeared as alternate light and dark 

 zones (Fig. 2), and the first regular increment was 

 formed the day after hatching (Table 2). The num- 

 ber of increments (NI) corresponded to the chrono- 

 logical age in d ( AGE ) of the larvae ( Fig. 3 ) . The slopes 

 of the regressions for haddock {NI-0.99AGE, /? = 148, 

 r~=0.96, P>0.0001) and pollock {NI=l.09AGE, n=9, 

 r''^0.82, P=0.0005 ) did not differ significantly from 1 

 (^test, P=0.637 for haddock and P=0.890 for pollock). 



Growth of haddock and pollock otoliths 



The check at hatching was clearly visible and incre- 

 ments were easily distinguished in the otoliths of 

 larvae sampled at sea (Fig. 4). The nuclear and check 

 diameters at hatching for haddock and pollock dif- 

 fered in different years (AN OVA, P<0.01) (Table 3). 

 For haddock, differences in nuclear check diameters 

 between 1991 and 1993 were not significant (AN OVA, 

 P>0.05). There were no significant differences among 



