Acoustical Behavior of Myripristis berndti — Salmon 
379 
any direction would be detected. Moray eels 
were prominent potential predators, often seen 
in pairs or larger aggregations in the same 
habitat as menpachi. 
Growl sounds produced in the laboratory 
were associated with more intense aggressive 
interactions. This sound is associated with ag- 
gressiveness between pairs of fish both willing 
to fight. In about half the observed cases these 
sounds followed nipping between the two fish. 
When one of the two fish fled, knocks were 
produced by the attacking fish and, possibly, 
also by the fleeing fish. 
In three of the four field recordings, there 
was no evidence of a crepuscular peak in the 
production of staccato and grunt sounds. In 
one recording, a dusk peak occurred (April 
28-29). This was the only case when the hy- 
drophone cable was not secured near the sur- 
face with an air-filled bottle. Movements of 
the loose cable on the bottom under the ledge, 
combined with decreased light intensities, may 
have been responsible for the production of 
these sounds. 
The response of laboratory populations to 
moray eels consisted of orientation to the eel’s 
head, investigation of its tail, increase in rate 
of swimming movements, and the production 
of many grunt and a few staccato sounds. The 
response of natural populations to a diver was 
similar acoustically, but the fish had room to 
escape by scattering to either side or back into 
darker recesses. More staccato sounds were pro- 
duced by laboratory populations during the 
first few seconds after the eel appeared, while 
grunts were produced throughout the 1 -minute 
recording period, though at a decreasing rate 
as time passed and the eel made no further 
movement after entering the cave. Apparently 
the tendency to produce grunt sounds habituates 
at a slower rate than staccatos. Probably staccatos 
represent the most intense warning response to 
danger stimuli. These sounds were also oc- 
casionally produced by startled menpachi during 
introductions of nonpredatory fish which sud- 
denly entered the cave. 
Sound-playback experiments to four labora- 
tory populations indicated that fish responded 
differently to various types of their own sounds. 
There was no observable change in the behav- 
ior of fish during playbacks of background 
sounds. Some fish oriented to the speaker when 
knocking sounds were played back, but did not 
move to the sound source. The response to 
playbacks of both staccato and grunt sounds 
involved immediate orientation, followed by 
movements toward the sound source. Playbacks 
of staccato sounds suppressed activity in H. 
rufus, i.e., the fish retreated into their cans 
during the playback, as would be expected 
when the territory also included a protective 
area. Orientation to the sound source occurred 
just outside the can and, in some cases, the 
fish moved toward and investigated the experi- 
mental speaker after the sound had been turned 
off. These differences in responses by both spe- 
cies to their warning sounds can be attributed 
to territoriality in H. rufus and its absence in 
menpachi. In both cases it is clear that M. 
berndti, and probably H. rufus, are capable of 
orienting to a sound source located a few me- 
ters away, and that staccatos (and grunts in 
menpachi) warn that a predator is present and 
also indicate his location. A warning sound 
with no directional information would be of 
limited use when large numbers of fish are 
aggregated in areas of low light intensity, prob- 
ably not alone sufficient to permit visual 
localization of a well camouflaged predator. 
Presumably, the responses in the laboratory are 
made to the "near field’’ components of the 
sounds, since they occur within a meter of the 
source. The results support van Bergeijk’s 
(1964) contention that fishes are capable of 
localizing sounds within the near field. It would 
seem that M. berndti , which shows such clear 
responses to some playbacks, would be a good 
species to test for sound localization at greater 
distances in the far field. 
Reproductive activities in fish have led to 
the evolution of one or, usually, two distinct 
types of sounds. One of these, usually produced 
by males, presumably attracts and/or sexually 
stimulates the female. Some examples are the 
"boat-whistles" of toadfish (Gray and Winn, 
1961; Winn, 1964), "purrs" of Notropis 
analostanus (Stout, 1963), and the sounds of 
male Bathygobius soporator and Chasmodes 
bosquianus (Tavolga, 1956, 1958). The same 
sound may function in aggressive interactions 
between males during the breeding season, as 
in the cod (Brawn, 1961), but often a second 
