For an initial approximation of tlie growth of 

 the juveniles, a linear regression was fitted to the 

 data with the hatching date set arbitrarily on 

 April 1. Then, assuming a hatching size of 3 nnn.,'' 

 I extrapolated the regression to obtain a better 

 estimate of tlie hatching date, which proved to be 

 in early May. The regression was tlien recalculated 

 with May 1 as the hatching date. The linear 

 growth is rei^resented by 



Z=0.3492-i8 + 0.104184T 



where / is standard lengili (cm.) and T is age in 

 days. Figure 2 shows the regression and the ob- 

 served data. Table 3 gives estimated leng^ths at 

 various ages. 



120 160 20O 



AGE (DAYS) 



FKiCHK 2. — Criiwth of juveiiik' alluuMire found in liillfisli 

 stomachs. 



Table 3. — Expected lengths {derived from the linear regres- 

 sion) at various ages of juvenile albacore from billfish 

 stomachs 



'The age is based on an estimated hatching date of May 1 and should be 

 considered only tentative. 



The data from this study have a l)earing on an 

 luiresolved problem in the biology of albacore — 

 that of determining the alisolute age of the fish. 

 Various investigators, using scales and vertebrae 

 for aging, have reported conflicting results in de- 

 termining the absolute age of albacore. Otliers at- 

 tempted to estimate the age of groups of albacore 

 tliat ajjpear regularly in the commercial catch. In- 

 vestigators now generally agree that the modal 

 groups in the Pacific albacore fisheries represent 

 year classes, but disagree on the ages assigned to 

 tlie year clas.ses (Clemens, 1961 ; Otsu and I'chida, 

 1963). 



The expected length that I estimated for 1-ycar- 

 old fish is most similar to that given by Aikawa 

 and Kato (1938), who made one of the earlier 

 studies on the age and growth of albacore. Their 

 results, however, were not generally accepted be- 

 cause they jiostulated rectilinear growth. Otsu 

 (1960) was not entirely satisfied witli the results 

 of fitting a Gompertz curve to his growth data, 

 which indicated that albacore required '■) years to 

 grow about 30 cm. TL ^ (total length). In a later 

 study Otsu and Uchida (1963) speculated that al- 

 bacore 30 to 35 cm. TL, which sometimes appear 

 in the commercial catches, were probably 1-ycar- 

 old fish, and albacore 50 to 55 cm. TL were 2-year 

 olds. Clemens (1961) and Bell (1962), on the other 

 liand, assigned lengths of 52 cm. and 57.3 cm. TL, 

 respectively, to 1 -year-old albacore. 



SEASONAL AND LENGTH DISTRIBUTION 

 OF JUVENILE ALBACORE 



Catch per unit of effort is often used to assess 

 apparent abundance of fishes. For this study I 

 used the number of juveniles per 100 billfishes as 

 an index of apparent abundance. Juvenile albacore 

 were found in stomachs of only tliree (.•striped mar- 

 lin, blue marlin, and shortbill spearfish) of the 

 five billfish species examined. In calculating the 

 monthly apparent abundance of juveniles, how- 

 ever, I used the total of all billfish species for each 

 month. I assumed that (1) no differences exist 

 among the billfish species in relative efficiency in 

 capturing prey and the apparent abundance of all 

 the l)illfish species did not greatly change over the 

 years and that (2) billfishes will feed on juvenile 

 albacore when they are available, not selecting 



' Matsumoto's (1958) observations suggest that tuna larvae 

 are about 2 to 3 mm. long at hatching. 



208 



' Total length as defined by Marr and Schaefer (1949). 



U.S. FISH AND WILDLIFE SERVICE 



