454 
Small pieces of material such as plastic tubing 
were rejected some time later. The time of 
retention was so variable that large fragments 
were always given for palatability tests. Caras - 
sins continued to suck at a large piece of 
Metapenaeus for up to 30 minutes. They sucked 
at a large piece of plastic tubing for up to 20 
seconds and then ignored it. 
P elates and Torquigener rapidly tore up 
large pieces of Metapenaeus. They bit at large 
pieces of plastic tubing once or twice and then 
ignored them. Small pieces were sometimes 
swallowed and ejected later, so large pieces 
were always used for tests of palatability. 
Any material allowed to remain on the floor 
of the tank was investigated from time to time 
by all fish. 
EXPERIMENT II 
COMPARISONS OF PALATABILITY 
Fragments of the tissues of a number of 
marine invertebrates were fed directly to fish 
in the comparison situation described above. 
Tissue of unknown palatability was compared 
with a known tissue. The initial standard was 
Metapenaeus , but when this proved to be more 
highly palatable than any other tissue tried, a 
less palatable substandard was adopted, Pyrazus 
ebeninus Brugiere (whelk). The palatability of 
tissue from 48 different species was tested in 
this way. Because of the subjectivity of the 
method, three levels of palatability have been 
defined on the basis of the reactions of each 
fish to different types of material. Each material 
has been assigned to a particular level, rather 
than placed on an absolute scale. 
Results 
Three levels of palatability have been defined 
on the basis of palatability comparisons: 
(1) Highly palatable (++) material was 
always eaten very rapidly by Pelates and 
Torquigener , with no preliminary sampling. 
Gambusia and Caras si us clustered about the 
material and continued to tear pieces from it 
for more than 10 minutes and up to 30 minutes. 
(2) Palatable (+) material was sampled for 
a short time before being eaten by Pelates and 
Torquigener. Gambusia and Carassius clustered 
about the material and tore pieces from it for 
PACIFIC SCIENCE, Vol. XX, October 1966 
a much shorter time, less than 10 minutes, and 
generally less than 5 minutes. 
(3) Unpalatable (0) material was not eaten, 
but the fish investigated it and often mouthed 
at it. Differences in degree of unpalatability 
were noticed: some material was swallowed and 
later ejected, some material was never swal- 
lowed. 
Table 2 lists the palatability of all materials 
tested in this experiment. Minor differences be- 
tween the four fish in the position of test 
materials on a nominal scale were not consid- 
ered significant. There were no differences in 
the assignment of materials tested to any level 
of palatability. 
Table 3 shows all materials, classed as palat- 
able and unpalatable, grouped with respect 
to colour. A chi-squared test (Siegel, 1956) 
shows that the differences in colour distribution 
between palatable and unpalatable are signifi- 
cant (i 2 = 16.57, P < .001). 
EXPERIMENT III 
COMPARISONS OF PALATABILITY WHEN TEXTURE 
AND CHARACTERISTIC APPEARANCE 
ARE DISGUISED 
In Experiment II, no examples of unpalata- 
bility were found where appearance was ob- 
viously responsible (since all materials were 
"tasted” before being rejected), but texture 
was believed to be of some importance. For 
these tests 13 unpalatable animals were selected 
to determine palatability after texture was dis- 
guised and general appearance was altered. 
A solution of gelatin in sea water was used 
as a base material to which test materials were 
added. A final concentration of 20% gelatin 
produced a gel stiff enough to be cut into small 
pieces which dissolved only slowly in water. 
The reaction of the fish to gelatin was the same 
as their reaction to plastic tubing. A homoge- 
nate of Metapenaeus was added in concentra- 
tions of 10-50% to render the whole palatable. 
A concentration of 10% (by weight) of 
Metapenaeus was sufficient to make the gelatin 
block highly palatable, and the reaction to 10% 
Metapenaeus was indistinguishable from the 
reaction to 50% Metapenaeus. 
Tissues for testing were stored in a deep 
freeze after collection, and ground in a blender. 
