Thompson et al : Age distribution and growth of Seriola dumerili 



369 



increase even after increases in fish length or weight 

 slows or ceases (Pawson, 1990). This finding provides 

 further evidence that otolith growth is more closely 

 associated with time than with fish size as the fish 

 becomes older. Secor and Dean ( 1989) suggested that 

 otolith growth was a function of both somatic growth 

 and time in young fish. We propose that as fish grow 

 and approach an asymptotic size, otolith growth con- 

 tinues at a basal level and does not reach an asymp- 

 tote. This is further evidence of the disassociation of 

 otolith growth and fish growth in older fish. Contin- 

 ued sagittal growth throughout the life of the fish is 

 the property that makes sagittae a useful ageing tool 

 (Gamboa, 1991; Casselman, 1990). 



Opaque zones (annuli) observed in transverse sec- 

 tions of greater amberjack sagittae were not as con- 

 sistent as those reported in other species such as the 

 sciaenids (Beckman, 1989; Casselman, 1990). Wil- 

 son et al. (1991) found that establishment of annual 

 patterns in transverse sections of sagittae from pe- 

 lagic fishes was best accomplished through interpre- 

 tation of a combination of growth features on the 

 surface of and in sectioned sagittae. Transverse sec- 

 tions of greater amberjack sagittae contained growth 

 features described by Smale and Punt ( 1991) for red 

 steenbras and by Wilson (1984) and Prince et al. 

 ( 1986) for billfishes. We used similar criteria for dis- 

 crimination of annual patterns in transverse sections 

 of greater amberjack sagittae. Our criteria included 

 external ridges on the lateral side of the sagittae, 

 ridges within the sulcus acousticus, internal opaque 

 and translucent zones, and the association of these 

 features with a common area where many growth 

 bands converged (Fig. 3A). All these growth features 

 were coincident with annuli that were validated by 

 using oxytetracychne-injected fish (Fig. 4). Smale and 

 Punt (1991) found similar patterns in transverse 

 sections of sagittae from red steenbras (Petrus 

 nipestris). Although they reported fish as old as 25 

 years, transverse sections of sagittae fi-om red steenbras 

 looked similar to those of greater ambeijack. 



Disagreement over age estimates based on the 

 above criteria was usually over an assignment of the 

 first annulus. In some specimens a prominent opaque 

 zone occurred adjacent to the core; we concluded it 

 was laid down during the first winter and therefore 

 did not consider it an annulus. Baxter ( 1960) reported 

 that the first annulus was formed on the scales of 

 yellowtail (as Seriola dorsalis) during their second 

 winter, which was 18 months after hatching. We con- 

 sidered the first annulus in amberjack to be that 

 which was laid down in the second year, when fish 

 were actually 15 to 21 months old. When corrected 

 for the discrepancy of the first annulus, the coeffi- 

 cient of variation and index of precision for the two 



readers were 0.15 and 0.11, indicating reasonable 

 consistency in annulus interpretation. Similar val- 

 ues were reported by Wilson et al. (1991) for Atlantic 

 blue marlin. 



Apparently the sex-related size difference in am- 

 berjack is due to age-related differential mortality. 

 Most of the greater amberjack collected were under 

 5 years old, although we found fish as old as 15 years. 

 Males ranged from 1 to 9 and females ranged from 1 

 to 15 years. Females represented 76% of the samples 

 above age 7. Burch ( 1979) found female greater am- 

 berjack to age 10 and males to age 8 for fish taken by 

 southern Florida charterboat fishermen. The only 

 other pelagic predator that demonstrates age-related 

 sexual dimorphism, where females outlive males, is 

 swordfish (Wilson and Dean, 1983). Greater amber- 

 jack also share a life history similar to both great 

 barracuda and cobia in that they are all moderately 

 long-lived pelagic reef species, and greater amber- 

 jack could be expected to exhibit similar patterns of 

 longevity for males and females. However, deSylva 

 (1963) found no age difference between sexes for great 

 barracuda (Sphyraena bar-racuda). Richards (1967) 

 reported no sex-related age differences for cobia 

 (Rachycentron canadum ) from mid-Atlantic waters, 

 and Thompson et al.^ reported similar findings for 

 cobia from the Gulf of Mexico. We conclude that 

 greater amberjack males die at a younger age than 

 do females. More research is needed to determine 

 whether these differences are real or are the result 

 of sampling biases used in our study. 



The relation between sagittal weight and age dem- 

 onstrated that otolith weight can be used to estimate 

 relative age. This finding is consistent with that for 

 other pelagic species (Wilson, 1984; Wilson et al., 

 1991 ). Sagittal weight can provide a useful manage- 

 ment tool because the random sampling and weigh- 

 ing of greater amberjack sagittae can be used to es- 

 timate age without the additional time and cost re- 

 quired for sectioning. We used a negative exponen- 

 tial to model the relationship between otolith size 

 and fish size. The fit of this model indicates that 

 otolith size continues to increase over time. 



Growth 



The von Bertalanffy growth parameter estimates 

 obtained in our study were similar to both Burch's 

 (1979) and Humphreys' (1986). Like Humphreys 

 (1986), we did not find differences by sex. Burch 

 ( 1979) reported L„ of 146.3 cm, 159.7 cm, and 164.3 

 cm for males, females, and sexes combined for 

 charterboat- caught greater amberjack in Florida 

 waters. Humphreys (1986) reported L^ of 149.3 for 

 Hawaiian greater amberjack (sexes combined). We 



