SKIPJACK IN HAWAII FISHERY 



297 



Size Composition 



Small skipjack are usually sought by the fisher- 

 men only wlien they are unable to locate larger 

 fish. One would expex't. therefore, that the occur- 

 rence of small fish in the catch would be invei-sely 

 related to the presence of the larger skipjack, and 

 this does seem to be the situation. The numbers of 

 large and small skipjack taken in the Oahu fishery 

 in each biweekly period of the 2 years under stud,v 

 is plotted in figure 13. In 28 biweekly periods 

 (1952, 8-20; 1953, 8-22) when large skipjack were 

 generally pi-esent, a tabulation was made to see 

 how frequently the changes in the number of one 

 size group were associated with similar or opposite 

 changes in the other. Opposite trends, e.g., the 

 number of large skipjack decreases from the pre- 

 ceding period while the number of small skipjack 

 increases, occur in 22 periods while similar trends 

 occur in 6. The probability of obtaining such a 

 distribution if the numbers of large and small fish 

 in the catch fluctuate independently of one another 

 is less than 0.01. 



Size composition appears to be important in tlie 

 determination of the general level of catch (fig. 

 14) and the pounds caught per unit-of -effort (fig. 

 15). Both increase rapidly with an increasing 

 proportion of large fish up to a ratio of 1 large fish 

 for 1 small fish. Above this ratio the total catch 

 continues to increase at a fairly rapid rate, but 



catch per unit of effort increases at a slower rate. 

 The likelihood of catches of large numbers of 

 indivduals seems to be loosely linked with the size 

 composition. During the interval included in this 

 study, the largest numbers of skipjack were taken 

 either when small fish were especially numerous 

 and very few large fish were available or when 

 large skipjack were in a majority (fig. 16) . When 

 small fish outnumber the large, but are less than 10 

 times as numerous, there seem to l>e factors that 

 work against the capture of a large number of 

 individuals. These factors, if they exist, are prob- 

 ably related to the distribution of the various size 

 groups in the population which supports the fish- 

 ery. A hypothesis concerning the structure of this 

 population is offered below. 



Conjecture 



In order to account for the variations in appar- 

 ent abundance of particular size groups in the 

 catches, it is necessarj' to hypothesize a skipjack 

 population consisting of at least three and pos- 

 sibly one additional element. In the approximate 

 order of their importance to the success of the 

 fishery in 1952-53. these are as follows: 



Group A : "Season fish," approximately 17-22 

 pounds in, weight, which Brock assumes to be in 

 either their second or third year of life. This 

 group is migratory. 



1 — I — I — r 



J 1 1 I I L_l 1 I I I I 



J I I I I I I 



BIWEEKLY PERIODS 



FiQUKE lii. — Estimated numbers of large and small skipjack taken in the Oahu fishery, 1952-53, by biweekly periods. 



