Sudekum et al .: Life history and ecology of Caranx ignobilis and Caranx melampygus 



505 



is much poorer. Over the range collected in both 

 studies, the relationship from Seki predicts W values 

 14.5-17% lower than our estimates; near L^, the dif- 

 ference is almost 18%. The larger difference compared 

 with that for C. ignobilis may be due to the small 

 sample size for one C. melampygus regression. 



Williams (1965) also fitted a length-weight expres- 

 sion based on 75 C. melampygus specimens from east 

 Africa. Converted to common units (SL in millimeters, 

 W in grams), his relationship is 



W = 1.4173 x 10" 5 SL 3 - 066 . 



Over the range of sizes occurring in both studies, this 

 expression predicts weights 10-17% lower than ours 

 and closer to those of Seki (1986b). Near L^ it is within 

 8% of our estimates, and for our smallest specimens 

 its estimates are about 27% lower. The large differ- 

 ences between predictions of Williams' model and ours 

 may reflect a difference in body proportions or in large- 

 scale "condition factor" between this species in the 

 NWHI and east Africa. Or the result may simply be 

 an artifact of sampling. Williams (1965) commented 

 that in his sample, males were probably larger on 

 average and more variable in weight than females. The 

 ratio of males to females was much larger in his sample 

 than in ours. 



Uchiyama et al. (1984) fitted length-weight data for 

 Seriola dumerili and Pseudocaranx dentex from the 

 NWHI to power functions. For all four of these large, 

 closely related species, the data fit a power function 

 model well and the parameter values are similar. 



Uchiyama and Tagami (1984) reported a preliminary 

 von Bertalanffy growth equation for Seriola dumerili 

 based on counts of daily otolith increments up to about 

 2 years of age (well below the inflection point of the 

 von Bertalanffy weight curve). These results show a 

 faster initial increase in size early in life in S. dume- 

 rili than in either Caranx species. The K values for 

 both Caranx species fall well within the range of com- 

 mon values for many large fishes. 



For both Caranx species, our counts of otolith growth 

 increments made with an SEM gave estimates of age 

 that could be fitted well to the von Bertalanffy age- 

 length model. The daily nature of increment deposition 

 was partly validated for C. melampygus by the tetra- 

 cycline marking experiment. Caution must be used in 

 extrapolating the one-to-one correspondence to fish 

 larger than those used in the validation experiment. 

 Recent work has shown that growth increments in the 

 otoliths of older fish may be deposited at intervals 

 greater than one day (Pannella 1971, Wild and Fore- 

 man 1980, Ralston and Miyamoto 1983). If so, age 

 would be underestimated and growth rate would be 

 overestimated. 



The close agreement between the growth rate mea- 

 sured for captive C. melampygus fed ad libitum and 

 the natural rate estimated from the von Bertalanffy 

 growth curve may be fortuitous. There are a number 

 of aspects of the captive situation that might be ex- 

 pected to create disparity between these results. Even 

 though the fish were fed at least daily to satiation, the 

 confined tank environment and human disturbance 

 may have affected their behavior or health and slowed 

 their growth. These fish are wide-ranging active pred- 

 ators, and the laboratory tank provides an unnatural 

 environment for feeding and metabolism. A limitation 

 of the captive feeding study was that only young, fast- 

 growing, sexually immature specimens were available. 

 It is not clear that the growth pattern or metabolic 

 parameters estimated from these experiments are fully 

 applicable over the life span of the fish. Relative growth 

 and ingestion rates are high at younger ages, and 

 otolith ring deposition is more likely to be daily. It 

 would be useful, but technically difficult, to perform 

 these experiments with large adults. 



The von Bertalanffy relationship obtained for C. ig- 

 nobilis predicts a value of SL^ = 1838 mm, and (by use 

 of our length-weight regression) W^ = 120.14kg. By 

 comparison, the largest C. ignobilis specimens reported 

 caught in Hawaii (with length estimates based on our 

 regression) are 86.71 kg (1648mm), 81.49kg (1613mm), 

 and two specimens of 68.10 kg reported as measuring 

 about 1524mm and 1626mm FL, respectively (about 

 1417mm and 1512mm SL, on the basis of our regres- 

 sions) (Chuck Johnston, Editor, Hawaii Fishing News, 

 Honolulu, pers. commun.). All these reports are from 

 the main, inhabited Hawaiian islands and reflect recent 

 conditions, following several decades of heavy fishing 

 pressure on the species. Larger specimens may occur 

 in the relatively unfished NWHI, where stocks are in 

 near virgin condition and sampling has been sparse. 



For C. melampygus, the predicted SL 0O = 897 mm 

 and W^ = 17.31kg. Catch records from the fishery are 

 not well documented. However, there are reliable 

 reports of a C. melampygus caught on Oahu that 

 weighed just over 19 kg and one caught on Lanai that 

 weighed about 13.6kg (Peter Dunn-Rankin, Univ. 

 Hawaii, Honolulu, pers. commun.). The largest speci- 

 men for which we found documented record was one 

 of our own: SL = 760 mm, W = 10.00kg. From sizable 

 collections by the Hawaii Division of Aquatic Resources 

 and by the National Marine Fisheries Service, the 

 largest specimens reported (with length estimates 

 based on our regression) were 8.20kg (698 mm) and 

 6.40 kg (642 mm), respectively. For both these jack 

 species, the values of L^ and W m derived from our 

 size-age data are close to or somewhat greater than 

 reported values. 



