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



497 



Energy budget and 

 population consumption 



Ingestion rates were measured in captivity. A group 

 of six C. melampygus and a group of four C. ignobilis 

 were held in separate sections of a tank of flowing 

 seawater 5.5m in diameter and 0.6m deep. Standard, 

 fork, and total lengths were measured for all fish, and 

 then they were allowed to acclimate to the tank for 1 

 month while being fed raw herring at least once a day. 

 Uneaten food was removed and the weight consumed 

 was calculated for each feeding. Maximum feeding 

 rates were estimated by feeding ad libitum, at least 

 three times daily during several intensive feeding 

 periods that ranged between 4 and 10 consecutive days 

 each. The fish used in this experiment were not the 

 same individuals used in the tetracycline marking 

 experiment. 



Age-specific rations (rates of food consumption) of 

 the two jack species in the wild and annual consump- 

 tion by their total populations were estimated by using 

 the measured ration in captivity for specimens of one 

 size, together with results obtained by the preceding 

 procedures for growth, reproductive output, and 

 length-weight relationship. Data for the captive speci- 

 mens were used in the basic energy budget model for 

 fish (Mann 1965 and 1969, Parrish 1975) to obtain the 

 respiratory metabolic rate, Q, 



where k 



C 



Q 



s 



kC = Q + S + G 



= ration assimilation coefficient, represent- 

 ing the fraction of the ingested ration 

 available for utilization in metabolic 

 processes, 

 = ration, or rate of food consumption, 

 = rate of respiratory metabolism, 

 = rate of production of reproductive 



material, 

 = growth rate. 



For the captive fish, ration (C) was measured. 

 Growth (G) was both measured and estimated from the 

 von Bertalanffy model. The value of k = 0.8 (Winberg 

 1956:209, Mann 1967 and 1969) was adopted. Since the 

 fish were prereproductive, the reproductive term (S) 

 was absent. The respiratory metabolic coefficient (a) 

 was estimated from the relationship 



Q = a W\ 



using the calculated Q, the weight (W) of experimen- 

 tal fish, and y = 0.8 as a reasonable approximation for 



most fishes (Winberg 1956:149, Mann 1965 and 1969, 

 Paloheimo and Dickie 1965 and 1966). This coefficient 

 was then used in the original model to calculate Q for 

 fish of any weight, W. No experimental results were 

 available to estimate a directly for C. ignobilis. In view 

 of the other similarities with C. melampygus and the 

 high probability of strong metabolic similarities, the 

 value of a for C. melampygus was also used for C. ig- 

 nobilis. The corresponding growth rate for any weight 

 was derived from the von Bertalanffy model (com- 

 puting age corresponding to weight from the model 

 directly and evaluating the first derivative of the model 

 at that weight). For fish larger than the SFR, the max- 

 imum observed GSI was used with the body weight 

 to estimate the rate of production of reproductive 

 material (S). With all three terms on the right side of 

 the model computed, ration (C) was readily determined 

 for fish of any size. 



The size-frequency distribution of the wild popula- 

 tion of each species was estimated by pooling length 

 and weight data for all specimens collected in this study 

 with data compiled by the Hawaii Division of Aquatic 

 Resources and the National Marine Fisheries Service: 

 in total, some 253 specimens of C. melampygus and 802 

 of C. ignobilis. Weight-class increments of 200 g were 

 used for C. melampygus and 500 g for C. ignobilis. The 

 fraction in each size-class was multiplied by the appro- 

 priate computed ration, and the results were summed 

 to estimate an individual ration representative of the 

 population as a whole. This ration of the representative 

 individual, multiplied by the population size for any 

 area, provides an estimate of consumption rate by the 

 entire population. 



Population sizes for both species were estimated for 

 one of the the major study areas, French Frigate 

 Shoals (FFS), using two methods. First, sightings of 

 the species made during 56 visual underwater transect 

 censuses in a variety of shallow-water habitats were 

 pooled. For a crude population estimate, for such wide- 

 ranging species, the distribution of these censuses over 

 the various habitat types was taken as representative 

 of all the habitats occupied by these fishes. Each sight- 

 ing was expressed as the number of fish seen per unit 

 area. Estimates of the total submerged area of FFS 

 of less than about 20 m depth (the apparent prime depth 

 range of these species locally) were based on data from 

 Atkinson and Grigg (1984), Agegian (1985), and J.J. 

 Polovina (NMFS Honolulu Lab., Southwest Fish. Sci. 

 Cent., pers. commun.). The product of this area and 

 the population density estimates from the visual cen- 

 sus provided rough estimates of the total populations 

 of both jack species at FFS. 



A second population estimate was based on the as- 

 sumption that the density of the almost unexploited 

 jack populations at FFS must be higher than the 



