FISHERY BULLETIN: VOL. 83, NO. 3 



subtropical waters, though juveniles formed a signi- 

 ficant component of the adult diet in the tropics. 



The absence of degenerating T. coryphaenae and 

 the positive correlation of parasite number and host 

 length suggest that the parasite was long-lived and 

 accumulated in the fish with age. The low numbers of 

 Tentacuiaria in the 57 -i- cm fish caught in New 

 Zealand indicate that these fish have had a different 

 history from their peers in the tropics. The bulk of 

 the skipjack tuna caught in New Zealand are 45 to 55 

 cm long. Less than 10% measure 60 cm or more 

 (Habib et al. 1980, 1981). We have concluded above 

 that the majority of New Zealand fish recently 

 arrived from the tropics. The T. coryphaenae data in- 

 dicate that the 57 -i- cm fish left the tropics at 45 to 

 55 cm long and have not returned. Evidently as fish 

 age, they become less migratory. This was hypothe- 

 sized by Kearney (1978). 



Large fish were not necessarily permanent resi- 

 dents in New Zealand, however. Of 17 57-t- cm fish 

 on which full dissections were carried out, 2 were 

 carrying the acanthocephalan R. terebra, a parasite 

 thought to be relatively short-lived (see above) and 

 not picked up in New Zealand. Raorhynchus terebra 

 was common in fish from Norfolk Island (area H). 

 Thus some of the large fish may have recently come 

 from areas as far away as Norfolk Island. 



The first two canonical variate analyses comparing 

 all areas sampled suggested that fish 40 to 57 cm 

 long had moved between areas and carried the 

 longer lived parasites with them. Parasitologically, 

 there was no evidence of more than one stock of skip- 

 jack tuna in the Pacific. Richardson (1983) observed 

 an east-west cline in the gene frequency of two en- 

 zymes across the Pacific. From an analysis of 200 

 gene frequencies he proposed an "isolation by 

 distance" model for skipjack tuna. In this, the degree 

 of mixing of skipjack tuna genes was inversely pro- 

 portional to the distance between the spawning 

 areas. Tagging data have confirmed that there is 

 some mixing of adult skipjack tuna in the central and 

 western Pacific (Kleiber and Kearney 1983), though 

 more than 95% of the tagged fish recovered during 

 the SPC program were caught within 1,000 mi of 

 their point of release (Kearney 1982). 



Schools of skipjack tuna have been observed to 

 break up when feeding (Forsberg 1980). This and 

 observations from aircraft where schools have been 

 seen to merge and later separate (Habib unpubl. 

 obs.) have led to the hypothesis that skipjack tuna do 

 not remain in a particular school for more than a day 

 or so. Certainly the pattern of recovery of SPC tags 

 suggested that tagged skipjack tuna underwent con- 

 siderable mixing amongst schools soon after release 



(Argue and Kearney 1983). However, using Mar- 

 quesas data we found that several parasites showed 

 strong school associations, particularly didymozoid 

 Nos. 19 and 20 (D. intestinomfiuscularis and Lageno- 

 cystislUnivitellanulocystis spp.). In another didymo- 

 zoid, Neometadidymozoon helwis from the gills of 

 Platycephalus fuscus, it takes up to a year for the 

 worms to migrate through the tissues, pair up, 

 mature, and die (Lester 1980). Though only a short 

 migration is needed for didymozoids 19 and 20, as 

 they are intestinal parasites, the worms are still like- 

 ly to be in the skipjack tuna for a period of weeks. 

 Thus, their strong association with particular schools 

 suggests that school half-life is likely to be in terms of 

 at least weeks rather than days. 



In New Zealand, the large school-school differ- 

 ences observed in the numbers of T. coryphaenae 

 and several other tropical parasites, especially in the 

 early arrivals, indicate that at the time of catching, 

 the New Zealand schools had not mixed sufficiently 

 to mask their previously distinct tropical faunas. 



Do schools remain intact for an extended period, 

 perhaps for the life of the fish? Sharp (1978) found 

 evidence of genetic similarity between individuals in 

 core schools, suggesting that some members of the 

 school were siblings. However, none of L. multi- 

 sacculatum, Coeliodidymocystis sp., or T. coinj- 

 phaenae, three long-lived parasites that showed 

 significant school-school differences in New Zealand, 

 showed any significant differences in the Marquesas. 

 This suggests that within the probable long life of 

 these parasites, fish caught in the Marquesas had 

 changed schools and had thus obscured any patchi- 

 ness in the distribution of the infective stages of the 

 parasites. The parasitological data, then, do not sup- 

 port the hypothesis that fish stay in the same school 

 for life. 



ACKNOWLEDGMENTS 



We are indebted to R. E. Kearney, A. W. Argue, 

 and other officers of the Skipjack Program, South 

 Pacific Commission, for much of our material, much 

 of the Tentacuiaria data, and assistance in preparing 

 this report. Other material was obtained with the 

 help of David Bateman, Heinz Tuna Cannery, Eden, 

 N.S.W.; Paul Dalzell, D.P.I. Fisheries Research, Ka- 

 vieng, P.N.G.; Bernie Fink, Van Camp Sea Food Co., 

 San Diego, CA; James Joseph, lATTC, La Jolla, CA; 

 Ted Morgardo, Star-Kist PNG Pty. Ltd., Rabaul, 

 P.N.G.; and Ronald Rinaldo, Southwest Fisheries 

 Center, La Jolla, CA. Their cooperation is greatly 

 appreciated. 



Taxonomic assistance was rendered by David 



354 



