940 



Fishery Bulletin 101(4) 



Materials and methods 



Pop-up satellite tags 



There has been rapid development in PSAT technology 

 over the past few years, and there are currently several 

 PSAT models commercially available from different ven- 

 dors. PSATs vary in many features: in the number of 

 functions (temperature, pressure, tag inclination, light 

 level) they measure; in the parameters and release time 

 for which they can be programed; in the onboard data that 

 they can manipulate and store, in the data they transmit, 

 in their emergency release mechanisms or emergency pro- 

 gramming (or both), and in their cost. PSATs have been 

 used on several large pelagic teleosts, including North 

 Atlantic bluefin tuna (e.g. Block et al., 1998; Lutcavage et 

 al., 1999), swordfish (Sedberry and Loefer, 2001), and blue 

 mariin (Graves et al., 2002). 



The two tag models used to evaluate postrelease survival 

 of blue mariin in this study — the Microwave Telemetry 

 Inc. (Columbia, MD) PTT-100 (5-day tag) and the Wildlife 

 Computers (Redmond, WA) PAT (30-day tag), are similar 

 in external appearance. Both are slightly buoyant, measure 

 approximately 38 cm by 4 cm diameter (including antenna) 

 and weigh between 65 and 75 g (air weight minus attach- 

 ment leader and tag head). The size of these tags is suf- 

 ficiently small as to not appear to impose a major drag on 

 a large marine teleost such as a blue mariin (Block et al., 

 1998). The greatest exterior differences between the two 

 tag types are their color (the 5-day tags are black and the 

 30-day tags are grey and white) and the presence of a small 

 metal emergency release mechanism on the attachment 

 leader of the 30-day tag. Both tag models can withstand 

 a minimum pressure equivalent to a depth of about 1000 

 meters, which is well below the maximum observed depth 

 of previous acoustic tracking analyses of blue mariin move- 

 ments (Holland et al., 1990; Block et al, 1992). 



The 5-day tags (n=7; described in this paper as 5D-1 

 through 5D-7) were programmed by the manufacturer to 

 detach from the fish 122 hours after activation. A release 

 time of five days was chosen for this PSAT because several 

 blue mariin tagged with conventional tags have been recap- 

 tured within five days, demonstrating a return to feeding 

 behavior (E. Prince, unpubl. data). Moreover, mortalities of 

 released blue mariin that have been observed with acoustic 

 telemetry have occurred within 48 hours of release (Pep- 

 perell and Davis, 1999). The 5-day tags measured water 

 temperature once an hour and stored the average of two 

 hourly values, for a total of 61 temperature means (from 

 122 measurements). 



The 5-day tags also reported two average inclinometer 

 values, one before the tag released from the fish and one af- 

 ter These values can be used to infer active forward move- 

 ment by the fish. This instrument measures the percentage 

 of deployment time that the tag was at an attitude of less 

 than 30° above horizontal by taking a reading every two 

 minutes and either adding or subtracting this percentage 

 from the running total value. Because this value is brack- 

 eted with maximum (255) and minimum (0) boundaries, 

 and measurements are taken at short-duration intervals. 



a mortality or tag shedding event displaying a lack of for- 

 ward movement will show clearly as an almost vertical 

 reading, even if the fish was initially quite active. 



The 30-day tags {n=2; described within this paper as 

 30D-1 and 30D-2) were programmed by the user to release 

 32 days following deployment, allowing for a full 30 days 

 of data collection. The release time for the 30-day tags was 

 chosen, in part, to test the assumption that five days was 

 a sufficient duration to capture the rate of postrelease 

 mortality resulting from interaction with longline gear, 

 and to obtain more detailed behavioral data over a longer 

 time interval. It should be noted, however, that a longer 

 release time may bias estimates of postrelease survival 

 because there is an increased period for tag malfunction, 

 physical damage to the tag, or other sources of mortality to 

 occur (Goodyear, 2002). Using the manufacturer's software 

 (PatHost programming software, version 2.06, Wildlife 

 Computers) the 30-day tag was programmed to record tem- 

 perature values ( sensitivity=0.05°C ) every minute, and the 

 data were collated for transmission as the fraction of time 

 during each hour-long period that the tag was within each 

 of 12 user-defined bins between the following temperatures 

 (°C): <5, 7.5, 10, 12.5, 15, 17.5, 20. 22.5. 25, 27.5, 30, and 60 

 (for 30D-1) and <7.5, 10, 12.5, 15, 17.5, 20, 22.5, 25, 27.5, 

 30, 32.5, and 60 (for 30D-2). Similarly, direct pressure mea- 

 surements (measured sensitivity=0.5 m) were taken every 

 minute and collated for transmission as the percentage of 

 time during each hour that the tag was within each of 12 

 user-defined categories between the following depths (m): 

 <0, 2.5, 5, 10, 15, 25, 50, 100, 250, 500, 750, and 1000 (30D- 

 1 and 30D-2). These two tags also recorded minimum and 

 maximum depths and temperatures for each hour during 

 deployment. Finally, the tags recorded light level measure- 

 ments every minute, and these data were used to calculate 

 a local time of midnight and duration of daylight, thereby 

 allowing a later estimate of daily position. 



Analyses 



The slightly buoyant tags detached from the fish after the 

 specified intervals, floated to the surface, and transmitted 

 archived data to satellites in the Argos satellite system. 

 Position information and sections of stored data were 

 captured with each satellite pass (Arnold and Dewar, 

 2001), transmitted to a ground station, and ultimately to 

 the authors by means of the Internet. Location data were 

 analyzed by using the computer program PROGRAM 

 INVERSE (version 2.0, National Geological Survey, 1975; 

 modified by M. Ortiz, NMFS Southeast Fisheries Science 

 Center, Miami, FL) to determine net direction and mini- 

 mum net displacement from the point at which the blue 

 mariin was tagged and released to the position of the tag 

 at the time of the first precise position (location code 1, 2, 

 or 3) determined by the Argos satellite system. Tempera- 

 ture data from the 5-day tags deployed off Florida were 

 categorized by time of collection as daylight, nighttime, 

 or a composite dawn and dusk period (one hour before 

 and after sunrise and one hour before and after sunset), 

 for analysis of temperature by light level using a one-way 

 ANOVA (Zar, 1999). 



