294 



Fishery Bulletin 103(2) 





Figure 1 



PIER umbrella dart used for external attachment of tags. 



Each style of dart was inserted through the midline of 

 the fish at the base of the second dorsal fin according 

 to the method of Block et al. (1998). 



Archival tags were surgically implanted either in the 

 dorsal musculature below the first dorsal fin (when fork 

 length was >110 cm) or into the peritoneal cavity (when 

 fork length was <110 cm). The dorsal musculature im- 

 plant was performed by making a 1-cm incision 3-5 

 cm below the first dorsal fin. A cold-sterilized trocar 

 (14 mm diameter) was then inserted into the muscle, 

 to a depth of 13-14 cm, within a plane parallel to the 

 pterygiophores but angled 45 degrees to the anterior. 

 The trocar was then removed and the tag was inserted 

 so that the light stalk was angled toward the tail. The 

 incision was then closed with a monocryl suture mate- 

 rial. This method was similar to that used by Musyl et 

 al. (2003). Interperitoneal implants were done according 

 to the method of Block et al. (1998). 



PSAT Tracker algorithm and analysis system 



We have developed an automated system, called the 

 PSAT Tracker Information System (PTIS), to improve 

 the accuracy and minimize the subjectivity and tedium 

 of matching data from different sources (tag and satel- 

 lite). It is an application of the Environmental Analysis 

 System (EASy) (System Science Applications, Redondo 

 Beach, CA) software that is specifically designed for 

 handling four-dimensional information (latitude, lon- 

 gitude, depth, and time). We describe the system in 

 terms of three processes; importing tag data and satel- 

 lite imagery, calculation of the optimal path of the tag, 

 and dynamic display of the path and associated tag 

 information. 



Importing tag data and setting parameters 



The PSAT tracker information system was designed 

 to support data formats of three tag manufacturers: 

 Wildlife Computers, Microwave Telemetry, and Lotek. 

 All three tag formats are imported into FIS and stored 

 in a universal relational database format for process- 

 ing. Key parameters used in the calculation of tracks 

 include time and position of tag deployment, time and 

 position of tag recovery, light-based estimates of lon- 

 gitude (provided by tag manufacturers), maximum 

 swimming speed of the tagged fish (estimated and 

 determined by the user), and a bracketed range of 

 latitude within which the program will search for SST 

 matches. Processing involves the temporal matching of 

 SST as recorded by the tag with that measured from 

 satellite imagery. It is important to note that the PTIS 

 user-defined latitude bracket is unrelated to the light- 

 based latitude estimates provided by the tag manufac- 

 turers; instead, it is simply a range set by the user to 

 include all possible movement of the animal during the 

 tag deployment. However, longitude estimates are tied 

 to the tag manufacturers' light-based estimates; the 

 user has the option of tying PTIS position estimates 

 directly to the light-based estimates or allowing the 

 algorithm to search a specified distance on either side 

 of the light-based estimate. 



For this study the maximum fish velocity was set at 

 4 knots. This was meant to be an inclusive rather than 

 an exclusive value, broadening the range PSAT Track- 

 er could search for SST matches. SST matches were 

 also constrained to remain within ±20 nautical miles 

 (±0.33°) of the manufacturers' light-based estimates of 

 longitude, based upon the observance by Hill and Braun 



