384 
Fishery Bulletin 99(2) 
The pattern of size distribution with depth is mirrored 
by the pattern of spawning and nonspawning with depth. 
Both smaller and nonspawning SBT are more abundant at 
depth, whereas both larger and spawning SBT are more 
abundant near the surface. The vertical distribution of 
SBT larvae suggests that SBT spawn at the surface (Da- 
vis et ah, 1990), as do caged Atlantic bluefin tuna (Th un- 
nus thynnus ) (Fushimi et ah, 1998). Surface-water tem- 
peratures on the spawning ground usually exceed 24°C 
(Yukinawa and Miyabe, 1984; Yukinawa and Koido, 1985; 
Yukinawa, 1987). These warm surface waters may be nec- 
essary for the survival of their eggs and larvae, but adult 
SBT normally feed in colder water (often as low as 5°C [Ol- 
son, 1980] ). Temperatures of 10°-15°C preferred by bigeye 
tuna (Hanamoto, 1986; Mohri et ah, 1996) may offer more 
favorable conditions for nonspawning SBT and explain 
their strong association with high BE indices on the 
spawning ground. 
Previous studies have shown that yellowfin tuna 
caught by purse seine and handline have higher go- 
nadosomatic indices than yellowfin caught by long- 
line (Hisada, 1973; Suzuki, 1988; Koido and Suzuki, 
1989). Histological studies have found that yellow- 
fin tuna catches from purse-seine sets and shallow 
(Taiwanese-style) longline sets have a higher pro- 
portion of actively spawning fish than catches from 
deep (Japanese-style) longline sets (Itano 7 ). Thus, 
spawning fish are more likely to be caught near the 
surface and nonspawning fish are more likely to be 
caught in deeper water. 
The biological basis for size partitioning with 
depth could be that large fish spawn more frequent- 
ly than small fish and, therefore, bigger fish will 
be caught at the surface more often than smaller 
ones. Spawning frequency is known to increase with 
size in female yellowfin tuna (Schaefer, 1998) but 
could not be determined for SBT. The pattern of size 
distribution may reflect recruitment into spawning. 
However, this hypothesis is unlikely because histo- 
logical examination of ovaries indicated that all SBT 
caught on the spawning ground were mature i.e. had 
advanced yolked oocytes (Farley and Davis, 1998), 
although this does not preclude the possibility that 
they might not be ready to spawn. The most likely 
reason for size partitioning is that the spawning fre- 
quency or the proportion of time spent spawning 
to time spent in a nonspawning condition increases 
with size. 
If the ability to tolerate higher than preferred wa- 
ter temperatures improved with fish size, then this 
would facilitate longer spawning episodes or more 
extensive feeding in shallow waters, both of which 
would produce the observed pattern of size distri- 
bution with depth. Although the ability to conserve 
heat in cold waters may increase with size in SBT, 
it is not clear what size-dependent processes might 
be involved in avoiding overheating at high ambient 
temperatures. 
We do not understand the temporal and spatial 
scale of vertical movements of SBT on the spawn- 
ing grounds in relation to spawning and feeding, nor 
Itano, D. G. 2000. The reproductive biology of yellow- 
fin tuna ( Thunnus albacares ) in Hawaiian waters 
and the western tropical Pacific Ocean: project sum- 
mary. SOEST (School of Ocean and Earth Science and 
Technology) 00-01, JIMAR (Joint Institute for Marine 
and Atmospheric Research) Contribution 00-328, 69 p. 
Univ. Hawaii, 1000 Pope Road, MSB 312, Honolulu, HI 
96822-2336, US. 
