Thomas and Moltschaniwskyi: Ontogenetic changes in size and shape of statoliths of Sepioleuth/s lessoniana 



641 



Additional variation in statolith shape described 

 on the second and third PCA axes was attributable 

 to differences among juvenile, rather than adult 

 squid (Fig. 6, A and B: Table 4). Growth of the sta- 

 tolith along the ventrolateral and lateral dome axes 

 and in total length were all important as statoliths 

 approached their final size and shape. The ventro- 

 lateral and total lengths were proportionally larger 

 in juveniles, whereas the dorsolateral length and to- 

 tal width were proportionally smaller immediately 

 before individual statoliths attained their final adult 

 shape. 



If the PCA score of each individual statolith on the 

 first axis, as a descriptor of shape, is used to esti- 

 mate age of the squid, then an improved correlation 

 between age and statolith shape is seen when com- 

 pared to the correlation between age and statolith 

 total length alone (juveniles 77 = 15, 7-=0.73: adults 

 77=61, 7-'=0.61)( Fig. 4B). There is an exponential rela- 

 tionship for both juvenile and adult squid but changes 

 in statolith shape slow in older individuals. Analysis 

 of adult and juvenile statoliths as separate gi'oups to 

 eliminate gross statolith size differences that were iden- 

 tified in the PCA only highlighted the overall variabil- 

 ity in statolith shape within the two age groups, par- 

 ticularly among adult squid. 



Discussion 



Results from this study suggest that neither statolith 

 total length nor weight measurements can be reli- 

 ably used to predict the age of a Sepioteuthis 

 lessoniana individual. Statolith total length is a rea- 

 sonable predictor of age for juvenile squid (less than 

 60 days old ) but cannot be used to predict age in older 

 individuals of this species. This conclusion is in con- 

 trast with that of Gonzalez etal. ( 1996) who observed 

 that statolith length in ///ex coindetti shows signifi- 



1.0- 



PCA 2(0.7%) 

 ventro-lateral 

 lateral dome 



all variables 

 PCA 1(98.6? 



1.0 



dorso- 

 .\ !lateral 



PCA 3(0.3% )0-l ; 

 lateral dome 

 dorso-lateral 



--:'.« 



-0.1 

 dorso- 

 lateral 



JSiL 





B 



PCA 2 

 •(0.7%) 



0.1 

 ventro-lateral 



lateral dome 



■0.1 



jwidth 

 ventro-lateral 



Figure 6 



Relative growth of the length of sta- 

 tolith dimensions. Principal component 

 scores for the length of five statolith 

 dimensions on the first three axes. The 

 percentages indicate how much varia- 

 tion in the data set has been described 

 by each principal component axis. Data 

 are logj„ transformed. Circles = adults; 

 squares = juveniles. 



cant dependence on age (males 77 = 170, 7-^=0.71; fe- 

 males 77 = 171, 7-=0. 74). However, they also suggested 

 that the use of statolith length for age estimations 

 would require verification over several years to con- 

 firm this relationship. Statolith length has also been 

 shown to reflect age in Illex illecebrosus females 

 (77=31, 7^'=0.96) (Morris and Aldrich, 1985), whereas 

 statolith weight provided a better reflection of age 

 (77=112, 7^^=0.88 ) than statolith total length in Todarodes 

 angolensis (Villanueva, 1992). Even combining all five 

 statolith dimensions to produce a description of sta- 

 tolith shape in this study provided only marginally 

 better age estimates for ■§. lessoniana individuals than 

 statolith total length or weight measurements alone. 

 Variation was evident in size between adult sta- 

 toliths in this study because the heaviest statoliths 

 did not necessarily possess the longest total lengths. 

 Additionally, as much as half of the variation in sta- 

 tolith total lengths was not attributable to age of the 



