644 
Fishery Bulletin 96(3), 1998 
the maximum size of skipjack tuna and other tunas 
that could be collected with smaller plankton nets 
was about 12 mm; the usual size was 3 to 5 mm. 
Thus skipjack tuna and other tunas larger than 10 
mm could easily escape from these gears. Davis et 
al. (1990) pointed out that net avoidance by larvae 
should be considered in calculating estimates of 
abundance of larval tunas. 
Sampling efforts to collect juveniles (larger than 
10 mm) have been reported in a few studies; King 
and Iversen ( 1962) tried to collect juvenile tunas with 
four kinds of trawl nets in the central Pacific, but 
only six juvenile tunas, size 18 to 60 mm, were col- 
lected. Higgins (1970) tried to collect juvenile tunas 
in Hawaiian waters using a midwater trawl net with 
a mouth opening 12 m x 8 m. He collected 578 skip- 
jack tuna and 417 other tunas. Most samples rarely 
had juvenile stages of skipjack tuna and other tu- 
nas. Takuno and Ueyanagi (1978) tried to capture 
juveniles with a small pelagic trawl net in the tropi- 
cal western Pacific, collecting 20 skipjack tuna and 
six yellowfin tuna, from 6 mm to 31 mm. These vari- 
ous results suggest that juvenile skipjack tuna and 
other tunas avoided the small trawl nets. The mouth 
opening and the towing speed of the Tansyu is much 
greater than those of small trawl nets used in previ- 
ous studies. We therefore expected that the size of speci- 
mens that would be collected would be larger than those 
collected in previous studies if the survey area and pe- 
riod were chosen appropriately. Indeed, the size range 
of skipjack tuna and other tunas collected with the 
Tansyu was much greater than that collected with 
smaller nets. In addition, large numbers of juvenile 
skipjack tuna and other tunas could be collected. 
Skipjack tuna and other tunas that reach the young 
stage are able to swim much faster than juveniles. 
In this study, the number of young skipjack tuna and 
other tunas collected was low, and these fish were 
caught only at night. These results indicate that net 
avoidance by young fish affected the number of speci- 
mens captured. The results of previous studies on 
swimming speeds of scombrids demonstrate that the 
“burst swimming speed” of adult tunas is usually 
from 10 to 20 fork lengths (FL) per second (Mag- 
nuson, 1978). If the burst swimming speed of young 
tunas (which might swim slower than adults) is, for 
discussion, assumed to be 10 FL per sec- 
ond, the speed at 20 cm FL would be 
estimated at approximately 3.9 knots. 
Therefore, the nets would need to be 
towed faster than 4 knots in order to 
sample young tunas of 20 cm FL. How- 
ever, few young skipjack tuna and other 
tunas were collected with the Tansyu 
with its maximum towing speed of 5 
knots. If large numbers of young skip- 
jack tuna and other tunas were able to be 
collected, we would learn more about the 
swimming speed of young life stages of 
these fish and also their ecological char- 
acteristics, for example their swimming 
behavior and vertical distribution. 
Other methods exist for collecting ju- 
venile and young tunas, i.e. by means 
of light traps (Thorrold, 1993) or from 
stomach contents of tunas and billfish. 
50 
0 20 40 60 80 100 120 140 160 180 
Standard length (mm) 
Figure 4 
Length-frequency distribution of Thunnus spp. collected by the midwater 
trawl net in 1992-94 trial surveys. 
Table 2 
Comparison of catches between daytime and nightime tows foriC pelamis and Thunnus spp. and total wet weight (g) of collections. 
K. pelamis 
Thunnus spp. 
Average catch 
(g) 
Occurrence Catch per tow 
(%) (inds.) 
Occurrence Catch per tow 
(%) (inds.) 
Day 
Night 
5405.5 
7010.1 
52.9 12.5 
48.1 8.1 
28.7 3.7 
27.0 3.0 
