348 
PACIFIC SCIENCE, VoL V, October, 1951 
month in a given area (see Tables 4 and 7) 
result in the diet for any period agreeing 
rather closely with the average daily diet 
throughout the year. Thus, an adequate 
sample for a single month, while appearing 
scanty for year-round appraisal, has much 
more value in this area than it would where 
extensive seasonal variation is the rule. Table 
1 indicates the number of fish examined each 
month for each locality. 
The stomach of each fish was removed and 
opened, and the contents were washed into a 
Syracuse watch glass. The organisms present 
were identified to the most exclusive taxo- 
nomic category practicable, then counted and 
recorded. 
For each area these calculations were made: 
the average percentage by number of the 
complete stomach contents which each type 
of organism constituted; the average number 
of organisms in the stomachs; and the per- 
centage of stomachs containing the organism. 
Emphasis is laid upon percentages by number 
rather than upon volume. While this method 
is not altogether satisfactory because the size 
of the different food elements varies con- 
siderably, it provides the best general idea of 
the relative importance of the various com- 
ponents because of the selective method of 
feeding. To circumvent misconceptions aris- 
ing from such treatment of data of this sort 
Hardy (1924: 19) , dealing with food analyses 
of herring, gave increased values to larger 
organisms to compensate for larger size. This 
practice has not been followed in this paper, 
so it must be borne in mind that ghost 
shrimps, hyperiid amphipods, crab megalopa. 
and shrimps are considerably larger than 
copepods and barnacle larvae. Since data are 
presented by areas, and since the size range 
among food organisms taken in each area is | 
not as great as in the food of some other 
species of fish, the percentages would not be 
greatly altered if compensation were made for 
variation in size of organisms. The only im- | 
portant difference in sizes occurs where cope- 
pod eggs are present in considerable quan- ! 
tity. The data for Kaneohe Bay and Pearl j 
Harbor should be considered with this dif- I 
ference in mind because adult copepods are 
many times larger than their eggs. 
In tabulating the data, the collections were 
grouped according to locality, and the gen- 
eral mean for an area was obtained by averag- 
ing the totals for the year. The results of the 
food analyses are summarized for Kaneohe | 
Bay in Table 2, for Ala Wai Canal in Table 5, 
for Honolulu Harbor in Table 8, for Pearl 
Harbor in Table 9, and for Hilo Bay in Table 
10. The relative importance (percentage by 
number) of the most important food organ- 
isms consumed in each area is diagrammed in 
Figure 1 so that food habits in the five areas 
may be compared. To assess objectively the 
extent to which nehu depend upon the 
various components of the plankton, an 
analysis was made of the gross aspects of the 
month to month fluctuation in plankton or- 
ganisms in Kaneohe Bay and Ala Wai Canal 
(Tables 3 and 6) . Possible seasonal change in 
food habits of nehu in these two baiting areas 
was determined by tabulating the food or- 
ganisms present in the stomachs month by 
month (Tables 4 and 7). 
TABLE 1 
The Number of Nehu Examined Each Month for Each Locality 
LOCALITY 
JAN. 
FEB. 
MAR. 
APR. 
MAY 
JUNE 
JULY 
AUG. 
SEPT. 
OCT. 
NOV. 
DEC. 
TOTAL 
Kaneohe Bay 
4 
8 
10 
— 
8 
— 
5 
12 
19 
6 
8 
— 
80 
Ala Wai Canal 
- 
1 
3 
- 
56 
— 
17 
- 
- 
4 
- 
- 
81 
Pearl Harbor 
2 
— 
— 
2 
— 
— 
— 
— 
- 
17 
- 
- 
21 
Honolulu Harbor. . 
— 
— 
— 
— 
— 
— 
20 
- 
- 
- 
- 
- 
20 
Hilo Bay 
- 
- 
- 
- 
- 
- 
- 
20 
- 
— 
— 
— 
20 
222 
