492 
PACIFIC SCIENCE, Vol. XXI, October 1967 
are not mutually exclusive. The observations 
provided less information about the other 
hypotheses mentioned earlier. All the above 
hypotheses consider the adaptive significance of 
floating material in the ecology of pelagic fishes. 
The stimuli that release the approach of fishes to 
the raft are not discussed. 
Protection from Predation 
At least nine species of fish, both large and 
small, reacted to the raft in a way that made 
them less vulnerable to predation. Typically, 
when a predator approached the raft, the prey 
formed a compact group very close to the under- 
structure. When the predator left or ceased 
harassments, the prey again dispersed about 
the raft. Often the predator chased the prey 
to the raft. The value of the raft to the prey 
was demonstrated by the fact that only one 
species, the amberjack, frequently caught fishes 
that had taken shelter under the raft. Observa- 
tions on individual prey species are described 
below. 
The most common resident, the freckled 
driftfish, usually took a position far below and 
downwind from the raft and was sometimes 
out of view. Driftfish were able to match their 
background. They had a silvery countershaded 
coloration when not under the raft, but took 
on a mottled brown coloration when close un- 
der it, and those collected from under an orange 
drogue buoy had an orange color. Most of their 
predator-avoidance activity was in response to 
dolphins, although some was in response to 
pompano dolphins, wahoos, bottlenose dolphins 
( Tursiops sp.), or to pilotfish which approached 
the raft swimming with a whitetip shark. The 
hundreds of such responses followed an un- 
varying sequence: when one of the predators 
came into the vicinity, the freckled driftfish 
suddenly formed a compact school and swam 
rapidly back to the raft or the parachute drogue. 
(They also fled to the raft when an observer 
entered the water.) When an amberjack was 
preying upon them, they remained within about 
20 cm of the viewing chamber. They attempted 
to stay on the opposite side of the chamber 
from the amberjack or dodged into the gaps 
between the frames of the viewing windows. 
When the amberjack was not actively feeding, 
the driftfish ranged out again. Small damselfish, 
pilotfish, greater amber jacks, and jacks behaved 
similarly to driftfish in response to predation, 
but did not change coloration. 
Rough triggerfish ranged far from the raft, 
sometimes out of sight. Their rapid return to 
it usually heralded the appearance of a predator 
(billfish, a great barracuda, bottlenose dolphin, 
whitetip shark) or apparent predators (schools 
of mackerel scad or a powerboat). They re- 
sumed ranging before the potential predator de- 
parted, except when the predator was a bottle- 
nose dolphin. None of the above species 
exhibited a predatory response towards rough 
triggerfish. The triggerfish did not return to the 
raft when manta rays appeared and they usually 
swam out and met approaching dolphins. Rough 
triggerfish and dolphins may often be associated 
in the absence of drifting material; sometimes 
they arrived simultaneously at the raft. 
On several occasions, the most successful 
piscivore, the amberjack, itself became the po- 
tential prey of dolphins and took shelter 
beneath the raft. Although amber jacks fre- 
quently ranged 10 to 15 m from the raft un- 
molested, when the dolphin began pursuit the 
amberjack eluded the predator by swimming 
close to the chamber. It remained there for 
some time before ranging out again. 
The dolphin, one of the largest residents, 
took shelter close under the raft three times: 
once in response to a bottlenose dolphin, once 
to a billfish, and once to a swimmer. Each time 
the dolphin swam around the chamber just un- 
der the flotation drums and took on a coloration 
(Fig. 5) that occurred in no other situation 
Fig. 5. The lower dolphin assumed the dark 
coloration when one of the observers entered the 
water. 
