766 



Fishery Bulletin 100(4) 



the day (Holland et al., 1990; Dagorn et al., 2000). They 

 are capable of exploiting cold, deep waters below the ther- 

 niocline because of their unique anatomy (Graham and 

 Dickson, 2001), which enables them to physiologically and 

 behaviorally thermoregulate by making vertical forays 

 into the warm mixed layer throughout the day to increase 

 their body temperatures (Holland et al., 1992, Holland and 

 Sibert, 1994). Apparent differences in behavioral patterns 

 of tunas around anchored FADs versus drifting FADs have 

 been discussed, but almost no empirical results have been 

 reported (Holland et al., 1990; Dagorn et al., 2000). 



Although conventional and sonic tagging studies, and 

 those on depth, temperature, and time of capture with 

 longline gear, have provided valuable information on big- 

 eye tuna biology and ecology, the present generation of ar- 

 chival (electronic data storage) tags can further our under- 

 standing of bigeye tuna movement, behavior, and habitat 

 selection. Archival tags record swimming depth, internal 

 and external temperatures, and ambient light levels. The 

 light data can be processed to provide daily estimates of 

 latitudes and longitudes (Hill, 1994; NMFS, 1994; Welch 

 and Eveson, 1999; Hill and Braun, 2001). 



This study presents results from analyses of data re- 

 corded by archival tags recovered from bigeye tuna that 

 were released in the equatorial EPO during April 2000. An 

 evaluation of the accuracy and precision of the geolocation 

 estimates derived from the archival tag data is provided. 

 Descriptions of movement and behavioral patterns, and 

 also habitat characteristics of bigeye, are presented. 



Materials and methods 



Tag releases 



Bigeye tuna were captured, tagged, and released at three 

 FADs in the equatorial EPO between 0°.52' and 2°01'N 

 and between 95°24' and 97''06'W, during the period of 

 15-22 April 2000. Tagging was conducted on the chartered 

 FV Her Grace, a 17.7-m, 99 gross-t. United States west- 

 coast-style live-bait pole-and-line vessel. The FADs were 

 taken out of the water before the vessel left the tagging 

 area in order to minimize the probability of short-term 

 recaptures, and to permit the fish to disperse. 



The archival tags used in this study were model Mk7 

 manufactured by Wildlife Computers (Redmond, WA) 

 (Wildlife Computers, 2002). The total weight of a tag in 

 air is about 32 g. The Mk7 tag is designed for implanta- 

 tion into the peritoneal cavity of the fish so that the sensor 

 stalk protrudes outside the fish through an incision in the 

 abdominal wall. A label, printed in Spanish, with informa- 

 tion about reporting the recovery of the tag and the as- 

 sociated reward (US$500) was encased in the epoxy of the 

 main body of the instrument. 



The depth, internal and ambient temperatures, and 

 light level were stored in the memory of the tag every 4 

 minutes. At this sampling rate, the memory of each tag 

 (2 MB) was capable of storing 3.8 years of data. The pres- 

 sure sensors measured over a range of -40 to 1000 m with 

 0.5-m resolution. The temperature sensors measured over 



a range of -40° to 60°C with 0.05°C resolution. The light 

 sensors had a logarithmic range from 3 x 10'" W.cm^^ 

 to 1 X 0^'- W.cm~^. Bright sunshine is approximately 2 x 

 10"^ W.cm^^. Light-level readings of about 200 and 50 are 

 observed in bright sunshine and in complete darkness, re- 

 spectively. Exploratory analyses of our data indicated that 

 the light sensor is able to identify dawn and dusk events 

 at 300 m. 



Each tag was rigorously tested and evaluated at sea 

 before deployment in bigeye tuna. Tags were programmed 

 to collect data from all four sensors at one-second inter- 

 vals, placed in a nylon mesh bag with a Sea-Bird SBE-39 

 temperature-depth probe, and sent down to approximately 

 300 m. After retrieval, the data were downloaded from 

 each tag and their performance was evaluated in relation 

 to the temperature-at-depth data collected by the Sea- 

 Bird SBE-39. 



Bigeye tuna specimens were captured by using handline 

 gear during the day and night. Each fish was brailed with a 

 heavy-gauge aluminum rigid-framed net of knotless web- 

 bing and landed on a wet foam pad covered with smooth 

 vinyl. The eyes of the fish were immediately covered with 

 a wet synthetic chamois, the hook was removed, and the 

 condition of the fish was determined. If the fish was in ex- 

 cellent condition (i.e. no damage to the eyes or gills and no 

 significant bleeding), the surgery required for implanting 

 the archival tag was initiated. An incision about 2 cm long 

 was made in the abdominal wall about 10 cm anterior of 

 the anus and about 2 cm to the left of the centerline of the 

 fish. Special care was taken to cut through the dermis only 

 and partially through the muscle, but not into the perito- 

 neal cavity. A gloved finger was inserted into the incision 

 and forced through the muscle into the peritoneal cavity. 

 The tag, sterilized by soaking it in Betadine solution, was 

 inserted, pointed end first, through the incision into the 

 peritoneal cavity. The tag was then manipulated to the 

 caudal end of the incision, and two sutures were placed 

 rostral of where the stalk protruded to close the incision 

 by using a sterile needle and suture material (Ethicon 

 (PDS II) size 0, cutting cp-1, 70 cm). 



Fish were also tagged with two serially numbered 12.5- 

 cm green plastic dart tags (Hallprint, Pty., Holden Hill, 

 Australia) by using tubular stainless steel applicators. 

 Tags were inserted into the dorsal musculature with the 

 barbed heads passing between the pterygiophores below 

 the base of the second dorsal fin, from both sides of the 

 fish. Information for reporting the recovery of the fish and 

 for receiving the reward (US$500) for the return of the fish 

 was printed in Spanish on these tags. Lastly, five to ten 

 mL of 100-mg/mL oxytetracycline hydrochloride solution 

 were injected into the dorsal musculature with a dispos- 

 able automatic vaccinator 



The 96 bigeye tuna released with archival tags were 

 measured to the nearest centimeter with a caliper 

 (mean=109.2 cm FL, 95% CI=1.9 cm, range=88-134 cm). 

 The fish were then picked up by hand and released back 

 into the sea by one or more individuals standing in a rack 

 at the stern of the vessel. The total time most of the fish 

 were out of the water was recorded to the nearest second 

 imean=2:17, 95% CI=0:06, range=l:27-4:19, n=86). All 



