1954; Yabe, 1951 ; and Yabe et a!., 1959). The sword- 

 fish larvae are easily recognized by their long snouts 

 and heavily pigmented elongate bodies. They have a 

 prominent supraorbital crest similar to that of the 

 marlins, but lack the enlarged posttemporal and 

 preopercular spines. Larvae above 8.0 mm are even 

 more distinctive; they have one or more rows of 

 spinous scales on each side of the dorsal and anal 

 fins, with those along the latter continuing forward to 

 the level of the pectoral fin. 



Although the important fishing areas for this 

 species are mainly in temperate waters, the larvae 

 and juveniles are found largely in tropical and sub- 

 tropical waters throughout most of the Pacific. Fig- 

 ure 5 shows the locations of captures of swordfish 

 larvae and juveniles below 80.0 mm recorded to date 

 and those taken by HL ships. A similar plot of cap- 

 tures, exclusive of those taken by HL, was pub- 

 lished by Jones and Kumaran (1964). (One capture 

 site at lat. 23°N, long. 174°W is plotted erroneously. 

 This should have been in the southern hemisphere.) 

 Our samples extend the distribution of young sword- 

 fish to waters east of the Hawaiian Islands in the 

 North Pacific, and partially fill in the gap between 

 long. 132° and 172°W in the equatorial and South 

 Pacific. The overall distribution, which extends 

 roughly two-thirds the breadth of the Pacific, is simi- 

 lar to that of blue marlin larvae. 



Although captures were spotty throughout the 

 western and central Pacific, there were enough to 

 show differences in spawning time in the various 

 parts of the Pacific. The probable month of spawning 

 (Fig. 5) was calculated for each individual, using the 

 growth estimate of 0.6 mm per day derived by Arata 

 (1954). According to these calculations spawning 



Table 3. — Summary of young swordfish (Xiphias gladius) 

 taken in plankton net tows in the Atlantic and Pacific 

 Oceans. 



Figure 5. — Localities of captures of young swordfish <80 

 mm SL in the Pacific. (The numerals next to each capture 

 site denote estimated month of spawning.) 



•Taning (1955) examined 60 larvae of which 53 were <20 mm; 



no breakdown of larvae <10 mm available. 



occurred in spring and summer (March through July) 

 in the central North Pacific and in spring (September 

 through December) in the western South Pacific 

 south of lat. 10°S. In equatorial waters between lat. 

 10°N and 10°S, spawning occurred in all months of 

 the year. Spawning also seemed to begin and end 1 or 

 2 mo earlier in the western Pacific in the Philippine- 

 Formosa area, as compared with the Hawaiian Is- 

 lands area. This is understandable when we con- 

 sider: ( 1 ) that post-larval swordfish are usually taken 

 in the Atlantic in waters having surface temperatures 

 above 23.5°C (Taning, 1955), (2) that in the western 

 Pacific this isotherm lies between Taiwan and the 

 Philippine Islands as early as February, and (3) that 

 in the central Pacific along the same latitude, the 

 23.5°C isotherm passes northward through the 

 Hawaiian Islands in March or April, a difference of 

 1 to 2 mo. 



A unique aspect about the captures of swordfish 

 young is that of the small numbers taken in plankton 

 nets, only 28.4% were larvae smaller than 10 mm 

 (Table 3). Among other pelagic fishes, such as spear- 

 fishes, tunas, mackerels, etc., most of the larvae 

 caught in plankton nets are below 10 mm. Perhaps 

 the proportion of larvae caught is reduced inordi- 

 nately by the disproportionate catches of juveniles. 

 Among other fishes, particularly tunas and mack- 

 erels, juveniles above 10 mm are rarely caught, ex- 

 cept in much larger gear such as midwater trawls. 

 The large percentage of juveniles up to 80-mm long 

 taken in plankton nets suggests that the swordfish 

 young either do not react to the net quickly enough to 



244 



