LESTER ET AL.: PARASITES OF SKIPJACK TUNA 



Table 5.— Comparison of within and between sctiooi vari- 

 ability in numbers of parasites per fish for two areas. 



'The probabilities that the ratio of the between and within 

 school variances is no greater than one. (Based on transform- 

 ed data, i.e., log (parasite no. + 1.0).) 



'The probabilities that the proportion of fish with more 

 than the area median is the same for all schools. 



3* * • = P< 0.001 ; * * = P < 0.01 ; * = P < 0.05; blank = P > 

 0.05. 



Rate of Mixing of Schools 



To estimate the rate of mixing of schools we need- 

 ed to know the distribution of the parasites among 

 schools before, and after, some known time interval. 

 This we did not have for any of the Marquesas 

 samples. 



In New Zealand, however, some approximate 

 calculations could be made because schools arrived 

 from the tropics at different times. Sixteen of the 19 

 east New Zealand schools were of similar-sized fish 

 and were all caught within 1 mo. These schools were 

 divided into two groups: "early arrivals" and "recent 

 arrivals". (This was done by ranking the schools 

 using a combination of four parasites whose 

 prevalences were positively correlated with each 

 other, Nos. 16, 18, 19, and 20, and which were 

 thought to be relatively short-lived parasites picked 

 up in the tropics. Thus high numbers indicated a 

 recently arrived school.) From catch data (Habib et 

 al. 1980), we calculated that there was an average of 

 3 to 4 wk between the capture of 25% and 75% of 

 the annual catch. This interval was taken as the ap- 

 proximate period between the arrival times of the 

 early group and the recent group. If mixing was 

 occurring, one would expect that the school-school 



differences for tropical parasites would be greater 

 when the fish first arrived (the recent arrivals) than 

 after they had been there for a few weeks (the early 

 arrivals). However, this we could not demonstrate. 

 Our sample sizes at this point were rather small 

 (eight schools in each category), and in fact the 

 reverse appeared to be the case, the early schools 

 having a generally higher variability than the recent 

 arrivals. This suggested that the early arrivals had 

 come from several areas (and still had not fully mix- 

 ed), whereas many of the later arrivals had perhaps 

 come from one area. 



DISCUSSION 



Ten of the 26 parasites counted were species of 

 didymozoid trematodes. These are almost exclusively 

 a tropical group. Yamaguti (1970), for example, 

 found 84 different species of didymozoid in fish 

 around Hawaii. None were recorded in checklists of 

 parasites from New Zealand (Hewitt and Hine 1972) 

 or Canada (Margolis and Arthur 1979). Thus, 

 although skipjack tuna are caught in both tropical 

 and temperate waters, their didymozoid infections 

 are evidently picked up primarily in the tropics. 



Larval didymozoids have been found in small fish 

 and in invertebrates. It is almost certain that the 

 definitive host becomes infected by feeding on an in- 

 fected intermediate host (Cable and Nahhas 1962; 

 Nikolaeva 1965). In the tropics skipjack 40 to 60 cm 

 in length feed largely on fish, squid, and stomatopods 

 (Argue et al. 1983). In New Zealand, however, they 

 feed almost exclusively on euphausids (Habib et al. 

 1980, 1981). This completely different diet in New 

 Zealand, together with the fact that no endemic New 

 Zealand fish are known to carry any didymozoids, 

 lead us to the conclusion that few, if any, 

 didymozoids are picked up in New Zealand waters. 



The occurrence of 10 species of didymozoids in 

 skipjack tuna caught in New Zealand, in numbers 

 very similar to fish of the same size caught in the 

 tropics, thus indicates that New Zealand and tropical 

 fish were found until recently in a similar tropical en- 

 vironment. Almost certainly, the New Zealand 

 fishery is based on fish that have recently migrated 

 from the tropics, and not on fish recruited as post- 

 larvae in temperate waters. This disagrees with tag- 

 ging data which show that the bulk of New Zealand 

 skipjack tuna of known origin were off New South 

 Wales 10 mo earlier. However, the tagging inference 

 is applicable to < 4% of the total New Zealand fish 

 (Argue and Kearney 1983). Our conclusion is in 

 agreement with Argue et al. (1983) who found no 

 juvenile skipjack tuna in the stomachs of adults from 



353 



