946 



Fishery Bulletin 101(4) 



be independent. It is therefore not possible to accurately 

 describe the effects (i.e. to reconstruct the actual data or 

 data ranges) of the averaging process of the 5-day tags. Al- 

 though we feel that any direct comparison of inferred 5-day 



tag data with the observed data from the 30-day tags would 

 be inappropriate because of the inability to describe this 

 averaging effect, we note that the reported 5-day tag data 

 are not inconsistent with those data from the 30-day tags. 



The 30-day tags provided considerable information on 

 blue marlin movements at depth. Each hour-long histo- 

 gram included both the maximum and minimum depths 

 for the hour interval, as well as the percentage of time 

 spent within each predetermined depth bin. For example, 

 during the hour between 10;00 and 11:00 a.m. on 25 July, 

 30D-1 reported a maximum depth of 28 m and a minimum 

 depth of m. During this hour the fish spent 42 minutes 

 between and 2.5 m, only 72 seconds between 3 and 15 m, 

 and almost 17 minutes between 15.5 and 28 m. 



Examination of the maximum and minimum depth val- 

 ues by hour suggests a relationship between movement to 

 depth and daylight hours (Fig. 4). Both blue marlin with 

 30-day tags made deeper dives during the day than at 

 night. This pattern is similar to diurnal vertical migra- 

 tions obsei"ved in bigeye tuna (Thunnus obesiis) (Holland 

 et al., 1990; Dagorn et al., 2000), and swordfish (Sedberry 

 and Loefer, 2001 ) and may relate to feeding within the deep 

 scattering layer 



Billfish survival 



Seven of the nine deployed tags successfully reported 

 data. Three independent lines of data transmitted by 

 these tags suggest that seven of the blue marlin car- 

 rying these tags and released from pelagic longline 

 gear survived for periods of five or thirty days. First, 

 changes in temperature (5-day tags) or temperature 

 and depth (30-day tags) were consistent with verti- 

 cal excursions within the water column. Secondly, the 

 direction and magnitude of net displacement cannot 

 be explained by local current patterns and thus imply 

 active movement. Finally, the inclinometer readings 

 (5-day tags only) are consistent with sufficient forward 

 propulsion to depress tags below 30° above horizontal 

 for almost one-half of the measurement periods. 



The thermal histories of three fish tagged with 5-day 

 tags (5D-1 off Bermuda and 5D-4 and 5D-5 off Florida) 

 showed Httle variation (Fig. 2, A, C, and D). Because 

 of the infrequent sampling of temperature by these 

 tags (once an hour, and two hourly values were aver- 

 aged), it is possible that short-duration dives to cooler 

 waters were not sampled, or that detailed information 

 was lost in the averaging of the two hourly measure- 

 ments. Alternatively, the thermal histories for these 

 tags could be interpreted as those offish that died and 

 remained floating at the surface. However, given the 

 preference of individuals for shallow depths recorded 

 by the 30-day tags, even the minimal variations in tem- 

 perature observed by the 5-day tags may be consistent 

 with movement to depths and postrelease survival. Ad- 

 ditionally, the net displacements cannot be attributed 

 solely to local currents (Fig. 1), and inclinometer values 

 indicate continued forward movement throughout the 

 five-day period. We maintain, therefore, that these fish 

 were alive for the whole five-day post-release period. 



