536 



Fishery Bulletin 101 (3) 



Materials and methods 



The archival tag used in the present study (Northwest 

 Marine Technology, Inc., Shaw Island, WA) had four sen- 

 sors — for external temperature, internal temperature, 

 pressure, and light intensity. The external and internal 

 temperature sensors had a 0.2°C resolution and response 

 times of three seconds and 20 seconds, respectively. Resolu- 

 tion of the pressure sensor was 1 m of depth between the 

 surface and 126 m, then 3 m down to the scale limit of 510 m. 

 The tags measured data every 128 seconds. 



Two types of data files were created within the tag 

 memory. One data file stored daily records containing date, 

 estimated times of sunrise and sunset, water temperatures 

 at m, plus two other selectable depths (we selected 60 m 

 and 120 m), and other information required or produced in 

 the course of location estimates for each day. This file is re- 

 ferred to as the "summary file" in this article, and it stored 

 daily data from the time the memory was last cleared. 



The second data file contained unprocessed time series 

 data records taken at 128-second intervals. The tag could 

 record at any integer multiple of its measurement interval 

 and a multiple of one was chosen. Each record consisted of 

 external temperature, internal temperature, pressure, and 

 light intensity, and corresponded to a known time. This file 

 is referred to as the "detail file" in this article. It can hold 

 about 54,000 records, or about 80 days of steady data at the 

 high rate chosen — a small fraction of the tags lifetime. The 

 time-series memory was divided into two sections, and the 

 size allocation between the two sections was determined 

 by the user The first section filled first and did not change 

 thereafter. The second section filled next, but once full, it is 

 then continually overwrote old data. Thus the first section 

 always contained the earliest data seen in a tag deploy- 

 ment and the second always contained the latest data. We 

 divided the detail file into two 40-day sections for releases 

 in 1995 and 1996, and into 20- and 60-day sections for 

 releases in 1997. Most of the analyses in this study were 

 conducted with the detail file. 



Prior to experiments on wild fish, we applied archival 

 tags to three pen-held Pacific bluefin tuna from 93 to 97 

 cm in fork length ( FL I at Kasasa in Kagoshima Prefecture 

 (31°25'N, 130°11'E) in November 1994 to observe the effect 

 of archival tag attachment and implantation on fish. One of 

 the fish that had an archival tag inserted in its abdominal 

 cavity was recovered 453 days after tag implantation, when 

 the fish was caught for sale in the market. 



Experiments on wild young bluefin tuna were conducted 

 near Tsushima at the northeastern end of the East China 

 Sea every November and December from 1995 to 1997. A to- 

 tal of 166 fish, ranging from 43 to 78 cm FL (age or 1 ), were 

 caught by chartered trolling vessels, tagged on the vessel by 

 inserting archival tags into their abdominal cavities, and re- 

 leased immediately. Details of the tag, its performance, and 

 the manner of tagging are described in Itoh et al. (2003). 



Thirty of the 166 archival tags (18.1%) were recovered. 

 The durations of the tags at sea were 50 days or less for 13 

 fish, 96-211 days for 13 fish, and 359-375 days for three 

 fish, all recaptured around Japan. One additional fish was 

 recaptured off the west coast of Mexico, on the east side of 



the Pacific Ocean, 610 days after release. Data could not be 

 downloaded from one archival tag released in 1995 and re- 

 covered 30 days after release; all other tags yielded data. 



Results 



Sample records of swimming depth, water temperature, 

 and temperature offish viscera as recorded in the detail file 

 are shown in Figure 1 for three days in winter and three 

 days in summer The fish changed swimming depth fre- 

 quently with rapid dives and ascents. The water tempera- 

 ture changed little in winter, but it changed frequently and 

 substantially corresponding to dives in summer. Visceral 

 temperature changed gradually. 



Diurnal and seasonal change of swimming depth 



Differences between daytime and nighttime swimming 

 depth, water temperature, visceral temperature, and the 

 temperature difference between water and fish viscera 

 (thermal excess) were examined by using the Mann- Whit- 

 ney Utest (P=0.05) with data in the detail file. Data taken 

 during the hour before and the hour after both dawn and 

 dusk (four hours in all) were excluded from the test in 

 order to distinguish clearly between daytime and night- 

 time. For this purpose, dawn and dusk were taken as the 

 times the tag sensed the first or final light of the day. The 

 average swimming depth was significantly deeper during 

 daytime for 70% of all recorded days, which accounts for 

 the additional observation that the water temperature was 

 significantly lower during daytime for 66% of the days. The 

 visceral temperature was significantly higher during day- 

 time for 71% of the days, and thermal excess was signifi- 

 cantly larger during daytime for 85% of the days. 



Fish spent about 40% of their time within a 0-9 m depth 

 range and the time spent within each depth interval de- 

 creased as depth increased. This concentration in the 0-9 m 

 depth range was observed at both daytime and nighttime, 

 but was more pronounced at night (Fig. 2). 



The vertical thermal profiles (Fig. 2) show the change of 

 depth range of the surface mixed layer, the ocean layer that 

 lies above the seasonal thermocline, for one year Although 

 young Pacific bluefin tuna aggregated in the 0-9 m depth 

 range for almost all months, swimming depth was more 

 broadly distributed in winter when the depth range of the 

 mixed layer was greater (e.g. January and March). When 

 the depth range of the mixed layer became less in summer 

 (e.g. May and July), fish tended to concentrate near the 

 sea surface. Then as the depth range of the mixed layer 

 became greater in autumn (e.g. September and November), 

 the vertical distribution of the tuna gradually expanded 

 toward deeper water. 



Vertical swimming behavior at dawn and dusk 



The fish commonly showed a distinctive vertical movement 

 at dawn and dusk. At dawn, after a slow and steady descent 

 for about 40 minutes to reach to a maximum depth of 82 

 ±28 m (average ±SD), fish suddenly and rapidly ascended 



