Itoh et al.: Migration patterns of Thunnus orientalts determined with archival tags 



527 



comparing measured water temperatures at three depths 

 to the water temperature maps at each depth. However, it 

 is quite difficult to obtain water temperature maps for the 

 whole range of times and areas where tuna migrate other 

 than those for SST. Therefore we decided for the latitudinal 

 estimation to rely on the SST maps and on longitude esti- 

 mated by the tags. One difficulty with the SST method for 

 latitude is that water of the observed surface temperature 

 might occur at two different latitudes, thus not implying 

 a single unique position. However because the latitude of 

 about 80% of all days could be uniquely determined from 

 SST, we considered the adjustment method taken here to 

 be acceptable for the purpose of the present study. Although 

 it was not possible to check the accuracy of the latitudinal 

 estimation independently, judging from the accuracy of the 

 longitude values we used to locate the appropriate North- 

 South stripe on the SST maps and from the rapidity of 

 temperature variations found along those stripes, which in 

 most cases tightly constrained our estimates, we expected 

 the accuracy of latitudinal estimation to be around one de- 

 gree, which would be sufficient for a study of Pacific bluefin 

 tuna migration. 



Some deterioration was observed in pressure sensors. 

 The need for recalibration of sensors after recovery should 

 be emphasized. 



Effect of the tag on fish 



Fish in this study were much smaller than those in other 

 archival tag studies of southern bluefin tuna and Atlantic 



bluefin tuna (Block et al., 2001; Gunn and Block, 2001). The 

 tagging success achieved confirms that the type of archival 

 tag we used can be applied to fish at least down to 43 cm 

 FL. No fish died because of the attachment of the tag during 

 the experiment on pen-held fish. The recovery rate ( 18.1% ) 

 for fish tagged with archival tags was similar to the rate 

 (19.1%) for those in the conventional tagging experiment 

 conducted in the 1980s off Nagasaki Prefecture, including 

 Tsushima, for the same size fish of the species (Bayliff et 

 al., 1991). This comparison should be made cautiously for 

 the following reasons. The unusual appearance of an archi- 

 val tag body would attract the attention of the finder who 

 gutted the fish and might lead to a higher reporting rate. 

 Increased fishing effort for young Pacific bluefin tuna in the 

 northern part of the East China Sea in the 1990s compared 

 to that in the 1980s might lead to a higher recapture rate. 

 The inconspicuous stalk of an archival tag which was the 

 only externally detectable sign of its existence might lead 

 to a low discovery rate. Indeed, because many recoveries of 

 archival tags were made by consumers while gutting the 

 fish, archival tags implanted in the body of the fish must 

 have been overlooked by fishermen and by sellers at fish 

 markets. However, judging not only by the similar but also 

 high recovery rates, it seems that damage and stress of 

 handling at implantation and that due to the archival tags 

 being carried by the fish did not have much more effect on 

 fish survival than did the conventional tags. 



Macroscopic observations of two wild fish recovered with 

 archival tags showed that the surgical injuries that occurred 

 during archival tag implantations healed after one month 



