376 
PACIFIC SCIENCE, VoL XI, October, 1957 
and a suggestion of a mode at approximately 
164 cm. This suggests a differential growth 
rate between the sexes. 
Wahoo show size variation associated with 
changes in latitude, as has been described for 
other scombrid fish. For example, Nakamura 
(1952: 101) found that the average weight of 
yellowfm tuna increased with latitude both 
to the north and south of the equator be- 
tween the west coast of Sumatra and the 
Nicobar Islands. With respect to wahoo, the 
relative abundance of the larger fish appears 
to decrease from Christmas Island in the 
south to Palmyra Island and Kingman Reef 
in the north. This is accompanied by an in- 
crease in the numbers of smaller fish from 
south to north, these trends appearing in both 
the male and female length distributions (Fig. 
6). The indications are that the wahoo are not 
distributed over the area in random fashion 
with respect to either size (Fig. 6) or sex 
(Fig. 2). Unless each area has a separate popu- 
lation with a different growth rate and sex ratio, 
there must be differential movement of indi- 
MALES FEMALES 
5 tt rrrm-n 1 n n m rrrrrrn 
3 ' n KINGMAI 
5 - N.I 97 Jh 
J l 
TTH T'TTTTT 
N REEF 
(I! 
n r 
n N .40 _ 
Sn 
D - 
5 _ N • 196 
a 
PALM' 
YRA 1. n 
J 
“ L N * 40 
n n 
3 - 
5 _ N • 107 
V" 
IGTON 1 . 
J 
N-II 7 _ 
N, : 
0 - 
N .44 • 
0 - 
!' U 
FANN 
\ 
IING 1 . 
A 
AJg 
0 - 
5 _ N - 114 
0 - 
CHRIS' 
1 1 1 1 1 1 1 U 1 1 1 1 1 1' 
fMAS 1 . 
rrrrm frr 
N -227 . 
1 1 1 1 1 1 1 M m 1 rfl H 
86 102 118 1^4 150 166 86 102 118 1^4 150 166 
89 K>5 121 137 1^3 169 89 K» 121 137 153 169 
LENGTH (CM.) 
Fig. 6. Length frequency distributions of wahoo by 
areas, March 1955 through February 1956. 
viduals or groups in order to maintain the 
observed size and sex gradients. This move- 
ment, however, need not necessarily be con- 
fined to the study area. 
This nonrandom distribution of the wahoo 
makes it almost impossible to study growth 
by the method of progression of dominant 
modes in size frequencies because of the great 
difficulty in obtaining representative samples. 
Therefore such studies were not attempted. 
LENGTH-WEIGHT RELATIONSHIP 
As no significant difference in the length- 
weight relationship was noted between sexes, 
the data were combined to calculate the fol- 
lowing regression equation which describes 
this relationship in the wahoo: Log weight = 
—9-4199 + 3.50583 log length, where weight 
is in pounds and length in millimeters. The 
mean weights for given lengths may be deter- 
mined from the smooth curve shown in 
Figure 7. This large exponent, 3.50583, re- 
flects the elongate, slender shape of the wahoo. 
As one would expect for a fish of this shape, 
a small increase in weight yields a relatively 
large increase in length. On the other hand 
LENGTH (FT.) 
Fig. 7. Length-weight relationship for wahoo from 
the Line Islands, February 1951 to June 1955. Regres- 
sion curve fitted by the equation: Log weight = 
— 9-4199 + 3.50583 log length. 
