BRANSTETTER ET AL.: AGE AND GROWTH OF TIGER SHARK 



summer (Clark and von Schmidt 1965; Dodrill 1977; 

 Branstetter 1981). Therefore, for simpHcity, a 1 

 June birthday was used to estimate actual ages. For 

 back calculations, ages were based on the age at the 

 formation of winter annuli; therefore, for summer 

 caught tiger sharks, there is a difference between 

 the actual age and the age at annulus formation (i.e., 

 a tiger shark taken in June that was aged at 6.0 

 years would be 5 + years of age in back calculations). 

 To estimate growth rates, observed age/length data 

 were applied to a computerized von Bertalanffy 

 growth model (Fabens 1965). Males and females 

 were taken in similar numbers, and both sexes for 

 both samples are represented graphically; however, 

 because of the small data base, sexes were combined 

 for all mathematical analyses. 



Apparent differences in mean lengths at age be- 

 tween the two samples were tested for significance 

 using ^tests (Snedacor and Cochran 1980), and 

 independent von Bertalanffy curves for the two 

 regional samples were tested for differences follow- 

 ing methods of Bernard (1981) using computer anal- 

 ysis (SAS Institute 1985). 



RESULTS 



A FL/TL plot of data from both samples (Fig. 1) 

 can be used for general conversions of lengths re- 

 ported in this paper. However, analyzed separate- 

 ly, the two regional samples had nonsignificantly dif- 

 ferent regression formulas for the relationships of 

 FL or PCL to TL: 



Gulf 



FL = 0.871(TL) - 13.5 



PCL = 0.788(TL) - 12.1 



Atlantic 



FL = 0.853(TL) - 10.1 



PCL = 0.797(TL) - 14.2 



{n = 33, r = 0.998) 

 {n = 34, r = 0.977) 



(n = 66, r = 0.994) 

 (n = 68, r = 0.992) 



Combining data for both samples, the relationship 

 of centrum radii (CR) to length (TL) (Fig. 2) could 

 be described by linear regression; 



TL = 14.72 CR + 51.15 {n = 64, r = 0.972) 



Although the regression did not pass through the 

 origin, no correction factor, such as the Fraser-Lee 

 method (Carlander 1969), was applied because this 

 factor did not adequately describe the rapid embry- 

 onic growth (Casey et al. 1985; Branstetter 1986). 

 For simplicity, this- isometric relationship was used 

 for back-calculating lengths at previous ages and did 

 not produce Lee's phenomenon (Table 1). However, 

 the relationship was slightly curvilinear and was 

 more accurately described by separating the data 

 into immature vs. mature specimens (< or > 310 

 cm): 



Immature - TL 

 Mature - TL 



17.7 CR + 20.18 

 (n = 44, r = 0.972) 

 7.6 CR + 190.21 

 {n = 20, r = 0.796). 



Figure 1.— Relationship between fork 

 length and total length for Galeocerdo 

 cuvieri taken in the Gulf of Mexico and off 

 the Virginia coast. 



Galeocerdo cuvieri 



FL = .860(TL) - 11.5 

 rr.996 

 n-99 



_] L 



_l L 



J I L 



100 200 300 400 



TOTAL LENGTH (TL) (cm) 



271 



