164 



Fishery Bulletin 97(1), 1999 



well (Ricker, 1975). The curve did not show the typi- 

 cal asymmetrical sigmoid curve but seemed to ap- 

 proach an upper asymptote (Gulland, 1983). It was 

 the method with the lowest relative standard errors 

 for all parameters. Although there is also consider- 

 able variation in mass at GR, we felt that the 

 Gompertz growth curve has merit, especially because 

 there is a large change in mass associated with a 

 small increase in length among large sharks. The 

 results of the MIR analysis could be interpreted such 

 that GR formation is not related to time of year but 

 to mass increase (some sharks taking a longer time 

 than others to gain the same amount of mass). 



Typically, opaque band deposition is associated 

 with summer growth (Cailliet et al., 1983b, 1986; 

 Kusher et al., 1992), and the nature of the last de- 

 posited band can be related to the month of capture 

 to verify this. Using method SC-A and only summer 

 and winter months, we found that the observed ra- 

 tio of translucent to opaque last bands did not differ 

 significantly from the expected ratio, assuming a 

 translucent band deposition in summer. When we 

 used nine ring counts, however, the analysis did not 

 show any relation between the nature of the last band 

 and season. The analysis could be considered statis- 

 tically weak owing to the low sample size but was 

 included to emphasize the accuracy of the last band 

 identification. In this study the band immediately 

 after the angle change was opaque in most verte- 

 brae, which is in keeping with Francis (1996) who 

 reported time of parturition of white sharks as spring 

 or summer. In addition, if our interpretation of ver- 

 tebral bands in shark BT433 is correct, the opaque 

 band would be formed in summer Because of the 

 inconclusive results of the centrum analyses of the 

 entire sample, however, more recaptures of sharks 

 injected with OTC are needed to confirm this theory. 



The relation between centrum diameter and shark 

 length was linear, as was found in C. carcharias by 

 Cailliet et al. ( 1985) and in several other shark species 

 (Cailliet et al., 1983b; Schwartz, 1983; Branstetter, 

 1987). For another species of the family Lamnidae, /. 

 oxyrinchus, Pratt and Casey (1983) found a slightly 

 curvilinear relationship for both females and males. 

 The relation between centrum diameter and mass 

 was multiplicative, which explains the slightly big- 

 ger difference in centrum diameter between the sexes 

 than in the relation between centrum diameter and 

 length. The differences between the sexes in both 

 relationships, however, are slight and are probably 

 not of biological significance. 



Back-calculated mass and length values were lower 

 than observed values. The differences between mean 

 observed length (or mass) and mean back-calculated 

 length (or mass) at each GR would decrease substan- 



tially if the observed GR were treated as GR 1. 

 This could be an indication that the angle change of 

 the corpus calcareum is not formed at birth but in 

 the first summer growth (Brown and Gruber, 1988; 

 Wintner and Cliff 1996). 



X-radiography to enhance the visibility of GRs in 

 elasmobranch vertebrae has been used successfully 

 on several species (Cailliet et al., 1983a, 1983b; Yudin 

 and Cailliet, 1990; Ferreira and Vooren, 1991). This 

 technique was also used in the only other ageing 

 study of C. carcharias by Cailliet et al. (1985). They 

 counted GRs directly from x-radiographs and used 

 silver nitrate staining to corroborate counts of larger 

 vertebrae that proved more difficult to read. In our 

 study, scanned images allowed for easy and rapid 

 counts of rings and measurements for back-calcula- 

 tions. In addition, scanned images were easier to in- 

 terpret because they showed less detail of the narrow 

 circuli and prebirth marks thcin did the x-radiographs. 



Prebirth marks in placental species are normally 

 attributed to the time of placenta formation and at- 

 tachment (Casey et al., 1985; Branstetter, 1987; 

 Branstetter and Stiles, 1987). In our study, prebirth 

 marks were found in C. carcharias vertebrae. No 

 comments on prebirth marks in C. carcharias, I. 

 oxyrinchus, or Alopias vulpiniis were made by Cailliet 

 et al. (1985), Pratt and Casey (1983), and Cailliet et 

 al. ( 1983a), respectively Branstetter et al. ( 1987 ) did 

 not find prebirth marks in Galeocerdo cuvier, another 

 aplacental species. Branstetter and Musick (1994), 

 however, found prebirth marks in their Carcharias 

 taiirus specimens and related the first consistent 

 prebirth ring to the size of the embryo when diges- 

 tion of the large quantities of eggs begins. We did not 

 relate prebirth marks to embryonic length using back- 

 calculations because they represent growth in utero. 



The APE indices for the three methods (5.3-6.1%) 

 were considered acceptable. They were lower than 

 those of the four methods used by Wintner and Cliff 

 ( 1996) for Carcharhinus limbatus (8.1-13.0%, «=80- 

 87) and those of Cailliet et al. (1990) who used "bow 

 tie" sections of Miistelus manazo (6.9-12.7%, n=28- 

 30). D-values (3.9-4.1%) were also considered to be 

 acceptable because they were similar to those of 

 Natanson and Kohler (1996) for C. obscurus (3,3%, 

 7i=42) and lower than those of Cailliet et al. (1990) 

 (6.8-12.7%, n=27-30). 



Age and growth estimates 



Only one shark injected with OTC was recaptured 

 (BT433). Although it was at liberty for an adequate 

 time period, an interpretation of the bands on the x- 

 radiograph was difficult, and therefore the results 

 were based mainly on viewing the vertebra with 



