90 



Fishery Bulletin 103(1) 



40°N r 



38°N 



36°N 



34° N 



32°N 



78°W 



75°W 



72°W 



69°W 



66°W 



Figure 3 



Minimum straight line distances traveled by a surviving white marlin iTetrapturus albi- 

 dus) (solid line) (A) and the drifting track of a transmitting tag (dotted line) in offshore 

 waters of the U.S. Mid-Atlantic Bight. The cross (B) denotes a moribund white marlin 

 that sank to the seafloor shortly after it was released, illustrating that dead fish did 

 not travel far from the initial tagging coordinates. 



fight time of 46 minutes and died 27 hours after tag- 

 ging, and DR02-04 had a fight time of 83 minutes and 

 died 64 hours after tagging (Fig. 4). There was no 

 significant difference in fight time (Z=0.4996, P=0.62) 

 between surviving and moribund white marlin, largely 

 due to the large range of fight times for moribund 

 animals. Handling times ranged from 1 to 5 minutes 

 per fish. 



Hook type had a highly significant effect on the 

 postrelease survival of white marlin (Fig. 5). Fish 

 caught on circle hooks experienced significantly higher 

 survival (20 of 20; 100%) than those caught on straight- 

 shank ("J") hooks (13 of 20; 65%) (Yates's corrected 

 CMH x 2 =7-386, P<0.007). There were also highly sig- 

 nificant differences in hooking locations and hook-in- 

 duced trauma between hook types (Fig. 5). Odds ratios 

 revealed that white marlin caught on straight-shank 

 ("J") hooks were 41 times more likely to be hooked 

 deeply (Yates's corrected CMH x 2 =H-48, P<0.001) and 

 over 15 times more likely to sustain hook-induced tissue 

 trauma resulting in bleeding (CMH x 2 =8-3, P<0.005) 

 than fish caught on circle hooks. Of the white marlin 

 caught on straight-shank ("J") hooks, half were hooked 

 in deep locations, and 70% of these fish were bleeding. 

 Four of the seven observed mortalities were those of 

 deep-hooked and bleeding fish. Overall, 56% of bleed- 

 ing, 40% of deep-hooked, and 57% of deep-hooked and 

 bleeding white marlin perished following release. In 

 contrast, all white marlin caught on circle hooks were 

 hooked in the jaw, and bleeding was evident only in a 

 single animal in which the hook point exited the edge 



of the eye socket but did not damage the eye. Addition- 

 ally, 20%' (8 of 40) of the white marlin in our study 

 became entangled in the line during the fight and were 

 "leadered" to the boat tail-first, a condition known as 

 "tailwrapped" (Holts and Bedford, 1990). This phenom- 

 enon was equally distributed with respect to hook type. 

 Five tailwrapped white marlin required resuscitation, 

 and two tailwrapped white marlin hooked in the jaw 

 with straight-shank ("J") hooks died. 



With the model developed by Goodyear (2002), the re- 

 sults of 10,000 simulated experiments at an underlying 

 true mortality rate of 35% indicated that approximate 

 95% confidence intervals for mortality estimates for an 

 experiment deploying 20 tags on white marlin caught 

 on straight-shank ("J") hooks range from 15%' to 59% in 

 the absence of confounding factors. A dramatic increase 

 in sample size would be required to improve the preci- 

 sion of mortality estimates (Fig 6). Doubling the sample 

 size (n=40) would decrease the 95% confidence intervals 

 to about ±15% of the true value and quadrupling the 

 number of tags (n = 80 PSATs) would reduce confidence 

 intervals to about ±10% of the true value. More than 

 200 PSATs would have to be deployed to lower the con- 

 fidence intervals to ±5% of the true value. 



The net displacement of released white marlin was 

 variable among individuals and across locations and 

 was used as an independent line of evidence to assess 

 survival. Surviving white marlin demonstrated move- 

 ment patterns that cannot be explained by surface cur- 

 rents alone. Distances and directions of displacement 

 are summarized in Table 2. White marlin tagged with 



