366 



Fishery Bulletin 97(2), 1999 



on a common point (Fig. 3A); these features were 

 usually associated with ridges on the lateral side of 

 the sagitta or associated with alternating opaque and 

 translucent zones, or with both ridges and zones (Fig. 

 3B). Similar structures were described and used to 

 age istiophorids by Prince et al. ( 1986) and Wilson et 

 al. (1991). 



Six of 48 oxytetracycline-injected fish were recov- 

 ered from the tag-recapture experiment. Oxytetra- 

 cycline marks were observed in sectioned sagittae 

 from all six fish (Fig. 4). The location of the oxytetra- 

 cycline mark in relation to presumed annuli sup- 

 ported the use of the gi-owth features described above. 



Figure 3 



Photomicrographs of sectioned sagittae from a 72-cm-FL (A) and 42-cm-FL (B) 

 north-central Gulf of Mexico greater amberjack. Photomicrograph A shows the 

 convergence zones (C) along the dorsolateral edge of the rostrum. Photomicro- 

 graph B shows the opaque and translucent zones in the rostrum; opaque zones 

 were interpreted as annuli (Al. 



The specimen in Figure 4 was captured in Septem- 

 ber and recaptured in March and had an annulus 

 consisting of a ridge and convergence zone between 

 the tetracycline mark and the margin. The combined 

 information obtained from six oxytetracycline- 

 marked otoliths confirmed that one annulus is formed 

 between November and March in the sagittae of two- 

 and three-year-old greater amberjack. 



Age structure 



Five hundred and ninety seven greater amberjack 

 sagittae were sectioned and aged by two readers. 

 Forty-five specimens were consid- 

 ered unreadable owing to poor 

 sample preparation. Pairwise com- 

 parisons between two readers in- 

 dicated that 9.6% of the counts dif- 

 fered by 1 year, 1.6% differed by 2 

 years, and 0.4% differed by more 

 than 2 years. Readers combined 

 had a coefficient of variation (CV) 

 of 0.15 and an index of precision 

 (D)ofO.ll. 



Age estimates of greater amber- 

 jack ranged from young-of-the- 

 year to 15 years. Mean age of 

 males (2.9 yiO was not significantly 

 different than that of females (3.2 

 yrXANOVA P>F=0.0641; however, 

 all fish over 9 years of age were 

 female. Most of the greater amber- 

 jack sampled were from one to 

 three years old (Fig. 5). 



A linear function best described 

 the relation between estimated age 

 and sagittal weight. The regres- 

 sion parameters were not signifi- 

 cantly different (ANCOVA) be- 

 tween males and females; there- 

 fore data were pooled and pro- 

 duced the relation 



Age = 0.15(SW) - 0.44, (r-^=0.75) 



where SW = sagittal weight (mg); 

 and 

 Age = amberjack age in 

 years. 



Growth was modeled by using 

 the von Bertalanffy growth model 

 (Fig. 6) to facilitate comparisons 

 with other values reported in the 

 literature. Our age data from 



