Crabtree et al.: Age and growth of Megalops atlanticus 



625 



Estimated age (years) 

 Figure 3 



Average observed lengths (± two standard deviations) and 

 predicted lengths from the von Bertalanffy growth model 

 for male and female tarpon, Megalops atlanticus. 



Otolith weight was significantly related to age (Fig. 

 4). The slopes of the otolith weight-age equations 

 (Table 2) were significantly different for males and 

 females (f-test, £=3.69, df=295, P<0.001). 



Discussion 



We obtained tarpon from a variety of fishery-inde- 

 pendent and fishery-dependent sources; conse- 

 quently, our sample was biased towards certain size 

 classes, and the size-frequency distribution of our 

 sample may not reflect that of the population. Most 

 small fish (<1,100 mm) came from fishery-indepen- 

 dent sources, and larger fish were sampled from the 

 recreational fishery. Our size distributions were bi- 

 modal and contained many small and large fish, but 

 only a few fish 900-1,200 mm in length because these 

 intermediate-size fish were too large to be sampled 

 effectively by our gear and were rarely harvested in 

 the recreational fishery. The size frequency of tar- 

 pon sampled from the recreational fishery was prob- 

 ably biased towards larger individuals. Most fish 

 were caught during tournaments or were kept as tro- 

 phies to be mounted by a taxidermist; presumably 



in both situations anglers selectively kept larger fish. 

 Sometimes tournaments imposed minimum size re- 

 quirements of as much as 50 kg on the fish harvested. 

 Because males were typically smaller than females 

 and rarely exceeded 45 kg, our samples from the rec- 

 reational fishery contained roughly twice as many 

 females as males, but this probably does not reflect 

 the population's sex ratio. Among the smaller tar- 

 pon (<1,100 mm) obtained from fishery-independent 

 sources, there were 79 males and 85 females and the 

 sex ratio was not significantly different from 1:1 

 ()C 2 =0.230, df=l,P=0.064). 



Age-validation experiments with OTC-marked 

 otoliths supported the hypothesis that tarpon otoliths 

 formed annual marks. Otoliths from 3 of the 18 OTC- 

 injected tarpon showed fewer than the expected num- 

 ber of increments. These fish showed relatively little 

 otolith growth following capture, and we were un- 

 able to resolve annuli that might have been present 

 on the otolith's margin. We could not read the otoliths 

 from three other tarpon and were unable to validate 

 the periodicity of annulus formation for these fish. 

 It is not surprising that several otoliths from OTC 

 experiments were rejected as unreadable because 

 almost 21% of otoliths from wild fish were unreadable; 



