DeMartini et a\ Maturity, size, and sex composition of XIphias gladius 



503 



Females 



40° 



35° 



0) 30° 



25° 



20° - 



I — I r 



- h 



- h 



-cr 



D 1 



I h 



I 1 I M 



. . I 



16 



69 



170 



32 



60 



66 



38 



155 



161 



472 



550 



319 



145 



34 



121 



117 



29 



6 



6 



31 



18 



Males 





ZZIZh 



I I- 



j^D- 



h-c: 



I — I r 



;] — I 



- h: 



H r 



h- rn I 1 



50 100 150 200 250 50 



Eye-to-fork lengtti (cm) 



100 



150 



200 



25 

 122 

 140 - 



41 



57 



48 



17 

 172 

 123 

 359 

 380 

 217 - 

 120 



55 

 212 

 184 



59 

 9 

 7 



46 



30 - 



250 



Figure 9 



Median body lengths lEFL, cm) plotted by 1-degree north latitude increments for female and male 

 swordfish (Xiphias gladius), caught by the Hawaii-based pelagic longline fishery during March 

 1994-June 1997. Also indicated are the respective 5th, 25th, 75th, and 95th percentiles, and sample 

 sizes (number offish; arrayed along right-hand abscissa). 



(first hypothesized as a "size-temperature mediated physi- 

 ological mechanism" by Hoey, 1991 ). 



Horizontal movements also incur different energetic costs, 

 as well as other types of costs, for fishes of different body 

 sizes (RofF, 1988), It is plausible that feeding and spawning 

 migrations involve different costs for individual swordfish 

 ranging over the broad spectrum of body sizes (90-210 cm 

 EFL, 13-180 kg; Fig. 8) commonly caught by the Hawaii- 

 based fishery. Because most small swordfish are males and 

 most large swordfish are females, it is further likely that the 

 costs of migration differ between the sexes as well as among 

 fish of varying body sizes that might migrate different dis- 

 tances. The significant relationships observed between lati- 

 tude and sex ratio and between latitude and body size for 

 swordfish of both sexes in the central North Pacific are con- 

 sistent with the latter hypothesis. 



Recurrent patterns of variation in the sex composition 

 of swordfish catches throughout the world's longline fish- 

 eries remain poorly understood. Observed patterns may 

 reflect natural differences in distributions of the sexes, 

 may represent consistent artifacts of different catchabil- 

 ities of the sexes during spawning periods, or both. A 

 challenging topic for future research will be the evalua- 

 tion of whether the differences reflect sexually distinct 



costs versus benefits for vertical or horizontal migration 

 that result from the disparate body masses of males and 

 females. 



Implications for stock assessment 



It is clear that stratifying swordfish catch statistics by sex 

 could appreciably reduce the variances of catch and effort 

 estimates, and future research should include simulation 

 studies that explore the quantitative effects of sex strati- 

 fication on stock assessments (Restrepo, 1991). Further, 

 if our observations for swordfish caught by the Hawaii- 

 based longline fishery have identified patterns that are 

 relevant to swordfish caught elsewhere, age-structured 

 stock assessments for swordfish in the Pacific should eval- 

 uate explicitly the potentially great importance of tempo- 

 ral and spatial dynamics in the sex and size composition 

 of catches. It will likely be necessary to stratify catches by 

 latitude of capture (and perhaps spawning and nonspawn- 

 ing period), in addition to standardizing CPUE statistics 

 for environmental variables and targeting (e.g. see He et 

 al., 1997; Hinton and Deriso, 1998; Bigelow et al., 1999), in 

 order to reduce the variances of abundance indices used in 

 stock assessments. 



