Acoustical Behavior of Myripristis berndti — Salmon 
377 
TABLE 7 
Response 
of Four Populations of 
Myripristis berndti 
to Sound 
Playbacks* 
sound off 
sound 
ON 
sound 
population 
LEFT 
right 
control 
exptl. 
PLAYBACK 
NUMBER 
SIDE 
SIDE 
side 
SIDE 
4 
34 
38 
22 
26 
Background 
5 
38 
46 
29 
27 
Noise 
6 
34 
50 
33 
23 
7 
35 
37 
22 
26 
Many 
4 
39 
33 
3 
45 
Staccato 
5 
39 
45 
11 
45 
Sounds 
6 
40 
44 
5 
51 
7 
27 
45 
3 
45 
Many 
4 
38 
34 
5 
43 
Grunt 
5 
40 
44 
16 
30 
Sounds 
6 
54 
30 
4 
52 
7 
39 
33 
6 
42 
Single Series 
4 
32 
40 
24 
24 
of Knocks 
5 
37 
47 
24 
32 
Single Staccato 
6 
36 
48 
20 
36 
Sound 
7 
33 
39 
7 
41 
* Values represent the total number of fish on each side of the tank every 15 seconds during minutes 1, 3, and 5 when 
no sounds were played back, and during minutes 2 and 4 when sounds were emitted from one (experimental) side of 
the cave. 
sounds produced by fish before and after one 
muscle was removed. The two bilateral mus- 
cles must then contract synchronously. The 
same results were obtained by Winn and Mar- 
shall (1963) with Holocentrus rufus. It may 
be that synchronous contractions of muscles 
associated with sound production are universal, 
but more evidence is needed. 
The relationship between the contraction rate 
of sound-producing muscles and the resultant 
frequencies of the sounds have been investi- 
gated electrophysiologically in a few fish. Po- 
tentials recorded from Myripristis berndti in 
this study and from H. rufus (Winn and Mar- 
shall, 1963) corresponded in temporal rela- 
tions to the pulses of sounds made by hand- 
held fish. Similar results have been obtained 
in the pigfish, Congiopodus leucopoecilis 
(Packard, I960), the sculpin, Myoxocephalus 
octodecimspinosus (Barber and Mowbray, 
1956), and for several species of catfishes 
(Tavolga, 1962). In squirrelfishes, handheld 
sounds contain frequencies from below 75 to 
about 4,800 cps. The fundamental frequency 
of the sounds (about 85 cps) is believed to 
be a direct translation of the muscle contrac- 
tion frequency (Tavolga, 1964), while the 
higher frequencies are harmonics resulting 
from resonance of the air bladder. It would 
be expected that removal of one sound-pro- 
ducing muscle would reduce the intensity of 
all frequencies, particularly the 75-150 cps 
octave band containing the fundamental, as 
was the case in M. berndti. 
Replacing some of the gas in the air bladder 
with water reduced sound intensities, and 
when all the gas was removed, no audible 
sounds were produced. The results indicated 
that the air bladder acted as a resonator in the 
production of sounds. Similar results were ob- 
tained with H. rufus (Winn and Marshall, 
1963) and other fishes in which an air bladder- 
muscle mechanism was involved in sound pro- 
duction (Tower, 1908; Hazlett and Winn, 
1962 ). 
Field observations during the day, 24-hour 
tape recordings, the behavior of populations in 
laboratory tanks, and locomotory patterns of 
single fish in the activity chamber lead to the 
following conclusions. Schools of menpachi 
congregate in areas of suitable cover during the 
day. Their presence can be detected during 
these times by the production of four distinct 
types of sounds. Fish can be caught by hook 
and line for a brief period after sunset (1930— 
2030 hours) as they emerge to leave the area. 
