Brodziak and Macy: Growth of Loligo pealei 



221 



and mature groups means were significantly differ- 



ent at the 5% level ( I Y,-Y M I =0.78 > MSD 



0.05U.M] 



=0.08) 



Because variances were heteroscedastic for the weight- 

 at-age data grouped by sex and by maturity stage, the 

 Games and Howell test was applied. The group means 

 of indeterminates and females ( I Yj-Y F \ =2.47 > 



MSD 



0.05U.F] 



=0.46), indeterminates and males (\Yj- 

 Y M I =3.08 > MSD 05|/ A/] =0.50 ), and females and males 

 ( I Y F -Y„ I =0.62 > MSD 005[FM] =0A0) were signifi- 

 cantly different at the 5% level. Similarly, the im- 

 mature and mature group means were significantly 

 different at the 5% level ( I Y T -Y M I =2.13 > 



MSD 



M ' 



0.05U.M] 



=0.32). Overall, the average growth rates 

 in length and weight were found to be significantly 

 different by sex and maturity stage. 



The pattern of growth of L. pealei analyzed by 

 month of hatching was also highly variable (Table 

 4). For indeterminate-sex squid, average growth in 

 ML ranged from 5.9 (June-hatched) to 11.2 (Septem- 

 ber-hatched) mm/month, whereas average growth in 

 weight ranged from 0.2 (July-hatched) to 1.9 (Sep- 

 tember-hatched) g/month. No seasonal pattern was 

 apparent in the average growth rate for indetermi- 

 nate-sex squid. However, few samples of indetermi- 

 nate-sex squid hatched during July to December were 

 available, and comparisons of growth rate by season 

 of hatching were not possible. For females, average 

 growth in ML ranged from 11.1 (December-hatched) 

 to 22.5 (October-hatched) mm/month whereas aver- 

 age growth in weight ranged from 2.2 (December- 

 hatched) to 19.5 (June-hatched) g/month. Average 

 growth rates were generally higher for squid hatched 

 from June to October when the average growth rate 

 in weight exceeded 10 g/month. For males, average 

 growth in ML ranged from 13.9 (December-hatched) 

 to 40.9 (June-hatched) mm/month whereas growth 

 in weight ranged from 3.4 (December-hatched) to 53.3 

 (June-hatched) g/month. Average growth rates of 

 males were generally highest for squid hatched from 

 June to October when the average growth rate in 

 weight exceeded 20 g/month. Overall, these data sug- 

 gested that the highest average growth rates were 

 achieved by adult squid hatched from June to October. 



The average growth rates by hatch month were 

 also tested for significant differences by using un- 

 planned multiple-comparisons procedures where the 

 natural logarithmic transformation was applied to 

 stabilize variance. The null hypothesis of homogeneous 

 variances for the samples grouped by hatch month was 

 rejected for growth in length (X 2 =44.41»X 2 005[11] 

 = 19.67) and in weight (X2=75.71»X 2 00 5[ii]=19.67). A 

 total of 16 pairs of group means for growth rate in 

 length were significantly different at the 5% level 

 (Table 5): July— January; July-February; July- 

 March; July-April; July-May; July-November; July- 



December; September-April; September-May; Sep- 

 tember-December; October-January; October- 

 March; October-April; October-May; October-No- 

 vember; and October-December. Overall, these tests 

 indicated that there were significant differences be- 

 tween growth rates of squid hatched during Novem- 

 ber-May and July-October. Similarly, for growth rate 

 in weight, a total of 16 pairs of group means for 

 growth rate in length were significantly different at 

 the 5% level (Table 6): July-January; July-Febru- 

 ary; July-March; July-April; July-May; July-No- 

 vember; July-December; August-May; September- 

 April; September-May; September-December; Octo- 

 ber-March; October-April; October-May; October- 

 November; and October-December. Overall, these 

 tests indicated that there were significant differences 

 between growth rates of squid hatched during No- 

 vember-May and July-October. 



There was no consistent pattern in growth of L. 

 pealei analyzed by year of hatching (Table 7). For 

 indeterminate-sex squid, average growth in ML was 

 lowest in 1990 (8.5 mm/month) and highest in 1992 

 ( 12.3 mm/month), whereas average growth in weight 

 was lowest in 1991 (0.7 g/month) and highest in 1992 

 (2.2 g/month). For females, average growth in ML 

 and weight was lowest in 1992 ( 16.3 mm/month and 

 7.2 g/month) and highest in 1990 (21.5 mm/month 

 and 15.2 g/month ). For males, average growth in ML 

 and weight was lowest in 1992 (21.9 mm/month and 

 13.2 g/month) and highest in 1991 (27.8 mm/month 

 and 25.0 g/month). Overall, these data suggested that 

 growth rates varied by sex among years. 



The average growth rates by hatch year were also 

 tested for significant differences by using unplanned 

 multiple-comparisons procedures where the natural 

 logarithmic transformation was applied to stabilize 

 variance. The null hypothesis of homogeneous vari- 

 ances for the samples grouped by hatch year was re- 

 jected for growth in length (X 2 =26.12»X 2 05[2] =5.99) 

 and weight (X 2 =47.90»X 2 005|2) =5.99). The group 

 means of 1990 and 1991 hatched squid were signifi- 

 cantly different in length ( I ^1990-^1991 1 =0.28 > 



MSD 



!] = 0.26) and weight ( I Y1990- 



0.05[1990,1991] 



y i99 i I =0.93 > MSD 5[i99o ,i99ij=0.67), but no general 

 trend in growth by year was apparent. 



Growth model 



The additive and the multiplicative error structures 

 were compared by testing whether the residuals of 

 estimated case-I growth curves were normally or log- 

 normally distributed. For 8 out of 10 subsets of 

 length-at-age data, the hypothesis of an additive 

 normal error structure was rejected at the 57c level. 

 In contrast, the hypothesis of a multiplicative log- 



