DeMartini et a\ Maturity, size, and sex composition o\ Xiphtas gladius 



497 



100 n 



0) 

 Q. 



170 220 



Eye-to-fork length (cm) 



Figure 5 



Nonlinear regression fits of the logistic model for percentage 

 sexually mature versus eye-to-fork length (5-cni EFL class), 

 weighted by square root of sample size per length class, for 

 male and female swordfish {Xiphias gladius) caught by the 

 Hawaii-based pelagic longline fishery during March 1994-June 

 1997. Solid and dashed best-fit lines are predicted relation- 

 ships, and analogous perpendicular lines are median body size 

 at sexual maturity (Ljf, intercepts), for females and males, 

 respectively. Symbols represent mean percentages mature per 

 5-cm EFL class. 



of day fished, and relatively more "tuna sets" (deep, day- 

 time soak) are made during the peak summertime tuna 

 season, particularly at latitudes at and below the Hawai- 

 ian Archipelago (He et al., 1997). We therefore reevaluated 

 the sex composition of swordfish catches south and north of 

 27°N using only the fish caught on sets on which swordfish 

 were targeted (He et al, 1997). A total of 5547 fish were 

 caught on swordfish sets, representing 84'^r of all swordfish 

 present in the parent sample. The male proportion of 585 

 swordfish caught on swordfish sets south of 27 'N was the 

 same (61%) as for all observer-sampled sets. Proportions of 

 females in catches at and north of 27 N also were the same 

 (56'7f ) for total and swordfish-targeted sets. 



Sex and size composition of the swordfish catch varied 

 with season (Table 6A). Small-bodied males and females 

 dominated lower-latitude catches during spawning as well 

 as nonspawning periods (Table 6Bi. Overall, about 64'^r by 

 numbers and 65'~'f by weight of all swordfish were caught 

 during March-July spawning periods. 



Several interesting spatial patterns were apparent in the 

 size composition data. Relatively greater numbers of "small- 

 bodied" (<25th length percentile; females: <126 cm; males: 

 <118 cm EFL) swordfish were caught south of 22-N (Table 

 6B). Relatively greater numbers of "large-bodied" (>75th 

 percentile; females: >172 cm; males: >156 cm EFL) fish were 

 caught north of 35°N (Table 6C I. Median body sizes of sword- 

 fish caught increased with latitude (regressions of ranked 

 median EFL on 1 bins, /•^'>0.8 for each sex, n=2\, P<0.001) 

 with the proportions of catch that were small-bodied (and 

 mostly male) greater south of about 22''N (Fig. 9). 



Discussion 



Sex-specific sizes at maturity 



Effects of methods Ninety-five percent confidence inter- 

 vals on our estimates of median body size (Lj^,,) at sexual 



