FISHERY BULLETIN: VOL. 85, NO 4 



the alternative that spawning is most intense 

 near the reef. 



Too few sailfish larvae were taken to make any 

 firm statements on distribution of larvae of differ- 

 ent sizes. However, they appeared to have a pat- 

 tern of size distribution with location similar to 

 that of black marlin larvae. 



The vertical distribution data show that, at 

 least during the day, preflexion larvae of blue 

 marlin and sailfish concentrate in the upper few 

 meters (perhaps upper 3 m) of the water column, 

 but not in the neuston. However, postflexion lar- 

 vae of blue marlin and black marlin are neu- 

 stonic. This ontogenetic vertical migration has 

 not been noted previously. The somewhat differ- 

 ent results from the limited interreef channel 

 samples could have been caused by turbulence 

 due to strong tidal currents in these narrow 

 passes. 



Using nonclosing nets, Ueyanagi (1964) stud- 

 ied vertical distribution of istiophorid larvae (all 

 taxa combined) over the upper 50 m and con- 

 cluded that during the day larvae were most often 

 caught at the surface and frequency of capture 

 decreased with depth. At night catches of larvae 

 were approximately evenly distributed over the 

 upper 50 m. More recent data (Ueyanagi unpubl. 

 data) confirmed this pattern for blue marlin, 

 striped marlin, spearfish, and sailfish larvae. 



It is possible that the observed horizontal dis- 

 tribution of istiophorid larvae in the Lizard Is- 

 land area results solely from a concentration of 

 spawning or at least hatching of eggs close to the 

 windward side of the reefs. However, it is likely 

 that additional factors are involved. The south- 

 east trade winds push surface water against the 

 windward sides of these reefs and although some 

 of the water flows across the reefs into the La- 

 goon, downwelling (anstau conditions) should 

 occur seaward of the reef. An organism which 

 maintains a position near the top of the water 

 column, as do the istiophorid larvae (or positively 

 bouyant fish eggs), would accumulate in such a 

 downwelling zone. A similar situation has been 

 described off the windward reef at Lizard Island 

 where larvae of a number of reef fishes with 

 shallow-living larvae were apparently retained 

 (Leis 1986). However, the istiophorid larvae ap- 

 parently disperse away from the surface at night 

 (Ueyanagi 1964) whereas the larvae retained off 

 windward Lizard Island tended to maintain their 

 day-time vertical distribution at night (Leis 

 1986). If they did leave the surface, the istio- 

 phorids might be advected away from the reef 



front. A further caveat against accepting the 

 "anstau hypothesis" as a full and simple explana- 

 tion for the distribution of istiophorid larvae in 

 the area involves the trade winds. During the 

 time the near-reef peak in istiophorid larvae was 

 best developed (2-5 November), the winds varied 

 from to 10 kt and from northeast to southeast 

 while on the other cruises, the wind was stronger 

 and varied from 10 to 30 kt and from east to 

 southeast. Finally, preliminary analysis of data 

 from the samples in which the istiophorids were 

 captured revealed that high abundances of a 

 number of reef fish larvae also occur off the wind- 

 ward reef face. Many of these were not near- 

 surface dwelling larvae. Further study of larval 

 fish distributions and their causes in this area is 

 clearly required. 



Whatever the causes for the distributions of the 

 istiophorid larvae very near the windward reef 

 face, it is somewhat surprising that the larvae of 

 epipelagic, oceanic fishes should be so abundant 

 in such a narrow band along the reefs. Sailfish are 

 known to spawn relatively close to land masses 

 rather than in the open ocean (Ueyanagi 1974c) 

 and black marlin are often found nearshore 

 (Nakamura 1985); blue marlin are truly oceanic 

 fishes (Nakamura 1985; Nishikawa et al. 1985). 

 Yet larvae of all three were concentrated in a 

 narrow band only 0.25 nmi (possibly to 1 nmi) off 

 the reef crest. If pelagic fishes such as istiophorids 

 concentrate their spawning very close to reefs or 

 if the larvae are retained there, it will be essen- 

 tial for such areas to be included in studies of the 

 larval biology of these fishes. The assumption 

 that open oceanic areas are the important nurs- 

 ery areas for epipelagic fishes seems at best ques- 

 tionable for istiophorids in the Coral Sea and sim- 

 ilar factors may apply to other taxa in this and 

 other areas. For example. Miller (1979) reported 

 much higher concentrations of yellowfin tuna lar- 

 vae, Thunnus albacares, in areas 200 m off the 

 Oahu shoreline than had been reported else- 

 where. 



Nearshore or near-reef areas may provide more 

 favourable habitats for fish larvae, including 

 those of many pelagic species, than do oceanic 

 areas. The larvae of jack mackerel, Trachurus 

 symmetricus , an epipelagic (albeit, neritic) fish, 

 are spread widely over oceanic and coastal areas 

 off California, yet larval mortality due to starva- 

 tion in oceanic areas can be much higher than in 

 coastal areas, presumably because of insufficient 

 concentrations of food offshore (Theilacker 1986). 

 This may apply to other pelagic fishes as well and 



764 



