40 
PACIFIC SCIENCE, Vol. IX, January, 1955 
Fig. 5. Length frequency distribution of larvae. 
Length Distribution of the Larvae 
The nehu larvae ranged in length from 
about 2 to 11 mm. It is presumed that larg- 
er larvae were able to escape the net. It 
was hoped that the progeny of pulses of 
spawning could be followed by the progres- 
sion of modes in the length distribution of 
successive samples of larvae. This was not 
possible because (a) the larvae were apparently 
dispersed from the sampling station too rap- 
idly and (b) the samples were not taken at 
sufficiently close time intervals. 
The composite length distribution of all 
larvae sampled at Station 4 is shown in Figure 
5. It is similar to that shown by Tester (1951: 
fig. 6) in that a main mode appears at 3 mm. 
and less pronounced modes appear at greater 
lengths. The latter differ slightly in position 
from those reported earlier but this may be 
attributed to the fact that sampling was con- 
fined to one station in the present investiga- 
tion and therefore did not adequately sample 
the larger larvae which drifted elsewhere. The 
minor modes are presumed to be real and to 
be related to the presence of "day groups." 
As in the previous investigation (Tester, 1951 : 
341), they suggest an average early growth 
rate of about 1.5 mm. per day. 
SUMMARY AND CONCLUSIONS 
1. Quantitative samples of nehu eggs and 
larvae were taken in replicate twice a week 
over a two year period at one station in Kane- 
ohe Bay, using a half-meter plankton net. 
The station was located at or near a focus of 
abundance of eggs in the southern sector of 
the bay. 
2. Spawning, as indicated by egg and larva 
catch, occurred erratically throughout the 
year, but with a summer maximum and a 
winter minimum. 
3. Variation in egg production between 
days could not be adequately explained by 
variation in temperature, chlorinity, or moon 
phase. 
4. Agglutinated eggs, with the embryo and 
yolk coagulated beyond recognition, formed 
a higher percentage than in a previous in- 
vestigation. Moreover in the present material, 
the percentage was higher in winter than in 
summer. The suggestion that agglutinated 
eggs were dead at the time of capture should 
be investigated further. 
5. Several possible explanations are ad- 
vanced for a large decrease in numbers be- 
tween the egg and larva stage, namely, drift 
of larvae from the sampling station, escape 
through the meshes of the net, sinking below 
the surface layers, and loss from mortality. 
A seasonally erratic egg to larva ratio is 
pointed out. 
6. The sampling is not adequate to trace 
pulses of spawning from the egg to the larva 
stages. A length frequency distribution of the 
larvae is included showing the presence of one 
major and several minor modes similar to 
those found in a previous investigation. 
REFERENCES 
Ahlstrom, E. H. 1953. Pilchard eggs and 
larvae and other fish larvae, Pacific Coast- 
1951. U. S. Fish and Wildlife Serv., Spec . Sci. 
Rpt ., Fisheries No. 102: 1-55. 
Barnes, H. 1952. The use of transformations 
