780 



Fishery Bulletin 101(4) 



Figure 1 



Otoliths of larval swordfish (Xiphiasgladius): (A) histological section through the braincase of a 21.0-mm larva; (B) left 

 sagitta from a 5.4-inni-PSL larva; (C) histological section of a right sagitta of a 33.3-mm-PSL larva; and (D) transverse 

 section of a right sagitta of a 47.5-mm-PSL larva ( C=core increment; l=first check; 2=second check; M=macula within 

 the otolith vestibule; 0=otolith; x's (B-D) mark intervals of measurement). 



larvae were collected, and back-calculating the geographic 

 origin of eggs with mean axial trajectories and velocities 

 of water currents for the Yucatan, Gulf Loop, Florida Cur- 

 rents, and the Gulf Stream (-1.5 m/s (Maul and Vukovich, 

 1993; Olson et al., 1994; Boicourt et al., 1998]) and the 

 Caribbean Sea (-0.2 m/s [Mooers and Maul, 1998]). 



Results 



Otolith structure and increment counts 



Sagittae and lapilli were round, extremely small, and 

 lacked rostra or sulci in larvae <5 mm PSL(Fig. 1, A and Bi. 

 A rostrum developed on sagittae at -5.5 mm PSL (Fig. IC). 

 Lapilli did not develop rostra, and remained symmetrical 

 with growth. 



Two checks, distinct zones of irregular increment spacing 

 and opacity, were evident on most sagittae (Fig. 1, B-D). 

 The first check was evident at the third increment on all 



sagittae examined. The second check was found on sagit- 

 tae from larvae >3.8 PSL but varied from the seventh to 

 tenth increment. 



Growth model 



A piece-wise regression (Table 1; Equations 5-7) with two 

 linear segments provided the best fit with biologically real- 

 istic parameters. An assigned intercept of 3.2 mm PSL was 

 used for the first segment; this value was obtained by adjust- 

 ing the length at hatching with the scale given by Yasuda 

 et al. ( 1978) and by accounting for shrinkage due to pres- 

 ervation. Growth rate for the first segment was 0.3 mm/d 

 and 5.9 mm/d PSL for the second segment (Fig. 2A). The 

 intersection of the two linear segments was at an estimated 

 age of 13.3 d AFH., 3 to 6 d after the observed second check. 

 The PSL of larvae at the intersection was 11.0 mm. 



Growth rate in LJFL, also modeled with a piecewise 

 regression, was 0.2 mm/d for the first segment (the upper 

 and lower jaws of larvae <11 mm PSL are of equal length. 



