Biology of the Aku — Brock 
103 
TABLE 4 
Comparison of Sex Ratios for Both Large and Small Fish for March-August 
AND September-December, 1949 and 1950 
LENGTH GREATER THAN 650 MILLIMETERS 
LENGTH LESS THAN 650 MILLIMETERS 
MARCH-AUGUST 
SEPT.-DEC. 
MARCH-AUGUST 
SEPT.-DEC. 
Males 
1,077 
304 
168 
465 
Females 
1,063 
208 
144 
322 
Total 
2,140 
512 
312 
787 
Probability 
.76 
<.0001 
.17 
<.0001 
is likewise a period of reduced availability, 
and, if the reduction in catch commonly ex- 
perienced during the fall and winter months 
arises from the departure of part of the popu- 
lation from the areas fished, it might be sug- 
gested that a disproportionate number of 
female fish are among those departing. 
The possibility that the observed changes 
in the sex ratio might be linked with changes 
in the sizes of fish available to the fishery was 
explored, as significant changes in the average 
size of fish in the catch does seem to occur 
during the year and some error may be ex- 
pected in sexing the smaller sizes. 
The fish sampled were placed in two groups 
by size— those larger than 65 centinleters and 
those smaller than 65 centimeters; however, 
the same pattern of sex ratios was displayed 
by both groups. (See Table 4.) If the length 
frequency curves (Fig. 1) are examined, it 
will be seen that fish of 65 centimeters are 
not abundant in the catch. This particular 
length, therefore, seems to be a natural point 
of division between large and small fish, at 
least for the dominant summer fishery, 
SUMMARY 
The length frequency measurements of Ha- 
waiian aku {Katsuwonus pelamis) for the period 
from the summer of 1946 to that of 1951 
contained definite modal groups which, when 
properly associated from summer to summer, 
indicated the probability of the existence of 
two and sometimes three year-classes in the 
landings which were possibly 1,2, and 3 years 
of age, respectively (fish referred to as 1 year 
old have probably reached or are about to 
reach the end of their first year and are, there- 
fore, beginning their second year of life), with 
modal lengths of 42 to 45 centimeters, 68 to 
73 centimeters, and 79 to 80 centimeters for 
each of the ages given above. 
A study of the schooling pattern of the 
aku indicated a strong tendency to school by 
size; no fish within a school, in the samples 
studied, differed by more than 20 centimeters, 
whereas fish present in the whole population 
sampled differed by as much as 50 centimeters. 
The possibility that the mechanism of this 
size segregation might depend upon the ex- 
istence of a higher swimming speed for the 
same proportionate effort for larger fish was 
suggested. An alternate hypothesis, that the 
schools were relatively stable aggregates, 
seemed less likely as the variances for these 
pure school samples were found to be sig- 
nificantly nonhomogeneous. 
The measurement of ova indicated that the 
spawning period probably extended from late 
February, March, or April to the first part of 
September, and the existence of several modes 
in the frequency distribution of ova diameters 
within individual ovaries indicated the likeli- 
hood of multiple spawnings. Fully ripe and 
spawned-out fish were both quite rare in the 
commercial catch during the spawning season, 
and, although these observations would in- 
dicate that fully ripe fish were not available 
