Frogs Learn to Avoid Eating Hairy Caterpillars
by David D. Olmsted (Copyright - 1998, 2006. Free to use for personal and
Last Revised August 31, 2006
Frogs generally do not show much in the way of adaptive behaviors yet as described
here they are capable of operant conditioning under the right circumstances.
Rana clamata (Medium) Following a Caterpillar (Schaeffer - 1911)
During the summer of 1911 Asa Schaeffer of the University of Tennessee collected
four different species of frog for observation in a cage. These were an adult
female Rana sylvatica (Wood Frog), a 3/4 grown Rana virescens of uncertain
sex, a medium 3/4 grown female Rana clamata (Green Frog), and a small
1/4 grown male Rana clamata. The frogs were acclimated to their cage (30 cm
by 30 cm by 40 cm) for 4 weeks before any observations were made. All ate
in a normal manner after four days except Rana virescens who only began to eat normally
after two weeks. The cage was made as "homelike" as possible with a large 25 cm
diameter dish located towards one end. It had a strip board board placed over it
to provide a sitting place. Dead leaves, grass and twigs were also placed
in the cage to provide hiding places. They were fed a variety of bugs at irregular
On July 29 Schaeffer placed 30 hairy caterpillars into the cage which
soon crawled over everything. The following is his description:
attempted to eat a caterpillar seven different times within an hour, but rejected
it each time. Following these trials no other caterpillars were visibly reacted
to. By attempting to eat a caterpillar and then rejecting it is meant this:
The frog shot out its tongue in the normal manner, bringing the caterpillar back
to the mouth, then extruding the tongue slowly, slightly wriggling it. In most cases
this muscular wriggling freed the caterpillar from the tongue; if it did not the
withdrawal of the tongue into the mouth scraped off the caterpillar in nearly every
case. Sometimes the tongue was wriggled without a previous withdrawal into the frog's
Rana clamata (medium) made only two trials of the caterpillars, rejecting
both, and then ceased to react toward them for a while. The frog then jumped into
the dish of water. Subsequently a caterpillar crawled along the rim of the dish
opposite the frog. The frog walked over to within 3 cm of the caterpillar and
paid close attention to its movements for about 15 seconds. Then the frog walked
backwards about 8 cm resuming the original somewhat indifferent attitude.
No further reactions toward the caterpillars were observed although they crawled
in most tempting situations.
Rana virescens made three trials of the caterpillars,
rejecting every one. No other response to the caterpillars was observed."
The next day (July 30) he colored six caterpillars red and six blue by sprinkling
powder on them. He then put these into the cage but the frogs still ignored
them. Later that day at 4:00 p.m. he recorded this:
"The caterpillars when disturbed
spin a thread of silk and suspend themselves on it until the disturbance is over,
when they crawl back again to their original position. A normal caterpillar
was observed to suspend itself in this manner about 5 cm in front of the Wood frog
(Rana sylvatica) and about 3 cm from the bottom of the cage. The caterpillar wriggled
and was snapped by the frog almost immediately. The caterpillar was held in
the mouth for a half-second. The tongue was then thrown out and held out for
two or three seconds. The caterpillar, however, stuck fast, and when the tongue
was withdrawn into the mouth the caterpillar went with it. It was soon swallowed
with seeming difficulty. The wood frog did not later react toward any of the crawling
Rana virescens. The caterpillars collect on the ceiling of the cage
and then remain quiet if left to themselves for a few hours. It thus happened
that the frog had not seen any crawling caterpillars for four and a half hours.
When I placed a caterpillar on the rim of the berry dish in which the frog was
sitting the caterpillar was snapped up at once but the tongue was quickly thrown
out, and the frog being in the water, the caterpillar was washed off. The
frog could not be induced to react toward another caterpillar.
(small) snapped up a caterpillar after it had dropped from the ceiling and
swallowed it with great difficulty. No visible attempt was made to reject
The above descriptions clearly show that the frog is capable of operant conditioning
type learning and that learning is quite quick. This learning also follows the classic
two stage adaptive strategy in which on the first day the frogs learned to
avoid the caterpillars but on the second day the frogs “forgot” that learning
and tested the association again but this time they learned to avoid them even more
quickly. None of the frogs responded to the caterpillars on the third day,
Yet hunger with its high level "food acquisition" motivation signal
can partly override the learning as described below: Here the motivation value despite
its attenuation by the adaptive weight is still large enough to trigger the motor
"August 9, 9:30 AM - The frogs have not been fed for two days. I placed
one of the hairy caterpillars into the cage. Each of the four frogs in turn
snapped up the caterpillar but in every case extruded the tongue subsequently
and shook the caterpillar off. Then the caterpillar was taken out of the cage.
12:30 PM - The caterpillar was placed into the cage again. Rana sylvatica
reacted first by making two short hops to orient so as to look directly at
the caterpillar. The head of the frog was then slowly lowered and brought
forward toward the caterpillar, but I could not see that the tongue was shot out
although I watched especially to see if this would happen. In a second or
two the head lurched forward a little more and then the tongue was very slowly
extended, barely touching the caterpillar. The tongue was now withdrawn and
then suddenly extruded with what appeared as a very slight attempt to shake
the caterpillar off. The caterpillar elicited no further response during the
next 45 minutes.
Rana clamata (small) and Rana virescens paid no attention whatever
to the caterpillar."
Rana clamata (medium) simply followed and observed the caterpillar
as it crawled up the ramp and up the side of the cage as shown in figure 1.
Position 6 and then 7 were taken after the caterpillar fell into the water.
This seems to show that the strategic behavior of following and orienting
is separate from the tactical behavior of snapping.
Shaeffer, Asa A. (1911) Habit Formation in Frogs. Journal of Animal Behavior.
Vol.1, No. 5, pages 309-355