Ecology of Four Apogonids — Strasburg 
conspicuous and in no way tend to camouflage 
the fish. 
A number of other fishes have been reported 
to inhabit sea urchin spines. Most of these bear 
contrasting stripes or bars which may match the 
spines in size and color. Such ecological situa- 
tions have been found elsewhere in the Apo- 
gonidae (Abel, 1960^:482, 1960^:34; Lachner, 
1955:43; and Magnus, in press), Gobiesocidae 
(Briggs, 1955:141; Mortensen, 1940:250; and 
Pfaff, 1942:413), Centriscidae (Davenport, 
1955:37; Herald, 1961:151; and Le Danois 
et al., 1957:121), and the sciaenid Eques, the 
clinid Malacoctenus aurolineatus, the gobiids 
Gobiosoma multi fas datum and G. ?70vemline- 
atum, and the gobiesocid Acyrtus rubiginosus 
(C. L. Smith, personal communication). 
Some spine dwellers match the urchin’s color 
but lack stripes. These include the apogonids 
Siphamia zaribae (Whitley, 1959:16), 5. versi- 
color (Eibl-Eibesfeldt, 1961:57), several spe- 
cies of the gobiesocid genus Arbaciosa (Jordan 
and Evermann, 1898:2340-2343), some speci- 
mens of the gobiesocid Acyrtus rubiginosus (C. 
L. Smith, personal communication), and the 
Hawaiian pomacentrid Dascyllus albi sella (R. 
A. Stevenson, personal communication). Of 
these, S. versicolor, at least, has a changeable 
color pattern which includes a striped phase. 
Peculiarly enough, its stripes appear only when 
the fish is away from its urchin host (Eibl- 
Eibesfeldt, 1961:58). 
Some of the authors mentioned above state 
that spine dwellers should align their stripes 
with the spines if their color pattern is to 
provide an effective camouflage. Apparently this 
is done by Paramia bipunctata, Aeoliscus, Dory- 
rhamphus, and possibly Diademichthys. These 
fishes are thus protected not only by the physical 
barrier of the spines, but also by the visual bar- 
rier of their camouflage. Such a dual safeguard 
may be necessary if their predators are capable 
of probing between the spines for food. 
touching upper edge of pupil, thence across midsides 
to caudal base; fourth from upper lip across iris, 
touching lower edge of pupil, thence across midsides 
to end of trunk; fifth from lower lip across suborbital 
to rear edge of opercle. Lines iridescent blue-white 
on head but faintly dusky on body. Tip of snout and 
lips sometimes red-orange. Iris yellow. Pectorals pale, 
other fins light red. 
339 
The other spine dwellers remain essentially 
horizontal, and their stripes are usually not 
aligned with spines. Obviously they are not 
camouflaged in this position, although the criss- 
cross pattern of lines thus presented may dis- 
rupt the fish’s outline. It is also possible that 
color and pattern are of little consequence once 
the fish is within the spine shelter. Certainly its 
brilliant yellow color is a blatant advertisement 
for A. novae guineae. 
A few observations were made on A. novae- 
guineae repopulation following departure of 
the fish from their urchin host. One urchin was 
removed from its cave and placed on open 
bottom 4 ft in front of the cave’s mouth. A 
ridge on the bottom prevented the urchin from 
being seen from the cave. As soon as their host 
was disturbed, the novae guineae left it and took 
shelter among the spines of a second urchin in 
the cave. The exposed urchin was watched for 
15 minutes but no fish left the cave to join it. 
The urchin was then lifted over the ridge and 
placed 3 ft in front of the cave’s mouth. In 2 
minutes a single novaeguineae swam to. the 
urchin and settled among its spines, followed by 
a second fish a few minutes later. The urchin 
crept to the cave’s mouth during the succeeding 
10 minutes, without being joined by additional 
fish. Two more novaeguineae joined it just as 
it crossed the cave’s threshold. The balance of 
its original population of fish remained with 
the second urchin during an additional 5 min- 
utes of observation. 
On June 18, 1955, an isolated coral mound 
containing four large Echinothrix, each bearing 
several A. novaeguineae, was heavily treated 
with rotenone and all of the novaeguineae col- 
lected. This mound was visited regularly for 
several weeks, and then less frequently. No 
novaeguineae had returned by July 12, but there 
were a number of them among the spines of 
one urchin on August 24. This indicates that 
novaeguineae is rather restrictive in its move- 
ments. Unfortunately the distance to the next 
aggregation was not recorded. 
Ten of the rotenoned specimens were exam- 
ined for food consumed. Of these, two fish 
were empty and three others (males) had their 
stomachs packed with embryonated eggs. Prob- 
ably these fish had been brooding eggs in their 
mouths and swallowed them under duress. Of 
