FISHERY BULLETIN: VOL. 8(i. NO. 4 



1986). (Henceforth in this report, to save space, both 

 sets and stops will be referred to as sets.) The former 

 results might be due merely to chance. The latter 

 results indicate that not all the fish remain together 

 at all times, but this information is of limited value. 

 A much more powerful method of analysis is needed. 

 Turner (1986) employed statistical tests to show that 

 tagged bluefin tuna caught during the calendar year 

 after release had mixed considerably with the un- 

 tagged population, but knowledge about shorter 

 term mixing is necessary for short-lived species, 

 such as skipjack tuna. The present report describes 

 what are believed to be new and useful methods of 

 analysis of the integrity of schools of fishes, using 

 data for tagged skipjack tuna released and recap- 

 tured in the eastern Pacific Ocean. 



MATERIALS AND METHODS 



The methods of tagging the fish are described by 

 Bayliff and Holland (1986). 



Tagged fish, or tags unaccompanied by the fish 

 to which they were attached, are recovered and 

 returned by fishermen, unloaders, and cannery 

 workers, accompanied by information which is used 

 to assign them to specific sets. Additional details 



regarding this aspect of the study are discussed later 

 in this report. 



The methods of collecting and processing the catch 

 statistics are discussed by Shimada and Schaefer 



(1956) and Joseph and Calkins (1969). Hennemuth 



(1957) and Shingu et al. (1974) described the 

 methods of sampling the fish and the calculations 

 employed to determine the size composition of tunas 

 in the catches. The areas shown in Hennemuth's 

 figure 1 have been changed several times since that 

 report was published, however; the areas used cur- 

 rently are shown by Peterson (1982: fig. 30). 



ANALYSES AND RESULTS 



If tagged fish released at the same location on the 

 same day mix thoroughly with the population of un- 

 tagged fish in the same area, schools of fish caught 

 in that area will have approximately equal ratios of 

 tagged to total fish, whereas if they do not mix 

 thoroughly some of the schools will have much 

 higher ratios than the others. In this report the 

 numbers of tagged fish recaptured in sets made at 

 various intervals after release are compared with 

 the numbers of tagged fish which would be expected 

 in those sets if the tagged fish had mixed thoroughly 

 with the rest of the population during the interval 



Table 1 .—Data used for analysis of integrity of schools of skipjack tuna. The ranges 

 of values of sets and average weight in pounds are explained in the text. 



632 



