Acanthurus triostegus sandvicensis — Randall 
223 
was captured between 3:20 and 4:25 A.M. and 
was empty. 
The digestive tracts of 57 manini in the 
acronurus stage (see section on transformation), 
which were obtained at night light stations in 
the Hawaiian Islands, Line Islands, and Phoenix 
Islands, were examined in the same way. Forty- 
five were completely empty. Four contained one 
to seven tiny fish scales; one of these also con- 
tained a shrimp chela. A few crustacean ap- 
pendages (mostly from shrimp) were found in 
three other fish, one of which had also eaten 
a barnacle metanauplius. The intestines of two 
fish had a small amount of unidentified chitinous 
remains. The crustacean appendages, metanaup- 
lius and fish scales were not free in the gut 
lumen but were caught between longitudinal 
folds of the gut. It seems evident that the acro- 
nuri do not feed at night. 
Four acronuri of Acanthurus were obtained 
from the stomachs of skipjack ( Katsuivonus 
pelamis) caught during the day in Hawaiian 
waters. Because of their being partially digested, 
I am not able to identify any of these acronuri 
to species; two, however, are too large to be 
manini. The stomachs and intestines of all four 
of these fish were filled with zooplankton. 
Shrimps and shrimp larvae were the principal 
food items. Also found were copepods, poly- 
chaetes, and the remains of larval fish. 
Food of Transforming Manini 
In an aquarium, manini in their first day of 
transformation from the acronurus to the juve- 
nile stage were not observed to feed. The di- 
gestive tracts of 30 specimens, collected in tide 
pools during the morning of their first day of 
transformation and preserved immediately, were 
devoid of food material. Of 24 similar speci- 
mens collected in the afternoon, 19 were empty, 
but 5 had eaten small amounts of fine filamen- 
tous algae and leptopel (for a discussion of 
leptopel see Fox, Isaacs, and Corcoran, 1951). 
Second-day transforming manini were seen to 
feed on algae in an aquarium, though not fre- 
quently, and preserved specimens of the same 
transformation age all contained small amounts 
of algae and leptopel. During the remaining 2 
or 3 days of transformation, feeding is progres- 
sively heavier. 
Food of Juvenile and Adult Manini 
The gut contents of juvenile and adult ma- 
nini consist almost entirely of relatively fine 
filamentous algae. The teeth of the manini are 
close-set and denticulate on the margins (Ran- 
dall, 195 6 A fig. 2a) like other species of Acan- 
thurus. They are therefore well adapted for feed- 
ing on filaments of algae. Experiments on feed- 
ing in aquaria showed that the absence of coarser 
algae in the gut contents is due to limitation in 
the size and strength of the jaws and teeth. Only 
the smaller branches of coarser algae, like species 
of Hypnea, were eaten. Small juvenile manini 
only ate the fine end branches whereas larger 
fish ate progressively larger branches. 
Inorganic sediment was rarely found in the 
stomach or intestines of manini. When branches 
of a delicate species of the red alga Polysiphonia 
were mixed with fine sand and offered to aquar- 
ium manini, the fish were most adept at pick- 
ing out the filaments which projected above the 
sand. The stomach of the manini is thin-walled, 
similar to that figured and described by Breder 
and Clark (1947: 295, fig. 1) for Acanthurus 
coeruleus. It contrasts sharply with the thick- 
walled, gizzard-like stomachs of some species of 
Acanthurus. Unlike the manini, the latter species 
normally ingest large amounts of inorganic sedi- 
ment with the algae upon which they feed. 
In addition to avoiding the ingestion of in- 
organic debris, juvenile and adult manini do 
not seem to feed on animal material (although 
some species of Acanthurus , such as A. xan- 
thopterus , will accept animal food readily) . Even 
when hungry, manini declined to eat any of 
several kinds of animal food which were of- 
fered. In an aquarium the fish were very skillful 
in avoiding the intake of small crustaceans and 
small masses of tunicates or sponges which were 
in close proximity to the algae on which they 
were feeding. Occasional small animals are found 
•in the gut contents, however. These are prob- 
ably taken in accidentally. In shallow water at 
Coconut Island, Oahu, an adult manini was ob- 
served feeding on algae close to an egg mass 
of the damselfish, Abudefduf abdominalis (the 
guarding fish had been frightened away ) . When 
the manini reached the edge of the egg mass, it 
ceased to feed, swam over the mass, and began 
feeding on algae on the other side. Other fishes, 
