780 



Fishery Bulletin 100(4) 



with their stalks. The batteries in four of the six tags, 

 which were in fish at liberty for 175 days or more, failed 

 and these tags stopped collecting data. Fortunately, how- 

 ever, previously collected data were preserved in the non- 

 volatile memories of the tags. Perhaps the most important 

 feature of archival tags is their ability to collect data on 

 the movement of tagged fish at frequent intervals from re- 

 lease until recapture (Hunter et al., 1986; Gunn and Block, 

 2001). There are, however, several factors that can affect 

 the accuracy of the geoposition estimates. These include, 

 but are not limited to, latitude, equinoxes, resolution of 

 the light sensor, light attenuation, and behavior of the fish 

 (Gunn and Block, 2001; Musyl et al., in press). 



Gunn et al.'* previously reported the accuracy of geolo- 

 cation estimates from the light data from archival tags 



Gunn, J. S., T. W. Polacheck, T. L. O. Davis, M. Sherlock, and A. 

 Betlehem. 1994. The development and use of archival tags 

 for studying the migration, behavior and physiology of southern 

 bluefin tuna, with an assessment of the potential for transfer 

 of the technology to groundfish research. In Proceedings of 

 ICES mini-symposium on fish migration, 23 p. International 

 Council for the Exploration of the Sea, Palaegade 2-4, DK-1261 

 Copenhagen K, Denmark. 



attached to southern bluefin tuna iThinuius maccoyii) 

 held in cages in the Indian Ocean to be about 0.5° in lon- 

 gitude and 1.5° in latitude. Welch and Eveson (1999) and 

 Musyl et al. (2001) estimated the accuracy of geolocation 

 estimates from the light-level data recorded by archival 

 tags by comparing the known and estimated locations of 

 tags that were attached to oceanographic buoys in the 

 north Pacific. The reported accuracy by Welch and Eveson 

 (1999) was ±0.9° in longitude and ±1.2° in latitude. The 

 reported accuracy by Musyl et al. (2001) ranged from 0,2° 

 to 0.3° in longitude and from 1.5° to 4.4° in latitude. In our 

 study, we estimated the accuracy of geolocation estimates 

 (longitude: 0.5°, latitude: 2.0°) by comparing the known 

 and estimated locations of 21 bigeye tuna on their days of 

 recapture (Table 3). 



We used our estimates of accuracy and precision as 

 criteria for filtering the daily geolocation estimates. In 

 other studies where archival tags were used to provide 

 estimated movement paths, geolocation estimates derived 

 from light data have been verified or adjusted by compar- 

 ing recorded temperatures from archival tags with maps 

 of estimated sea-surface temperatures from satellite data 

 (Gunn and Block, 2001). In the equatorial EPO, sea-sur- 



