A group of Italian animal psychologists from the universities of Padua, Trento and the Civic Museum of Rovereto experimentally proved that frogs can distinguish between small numbers, such as one, two and three but not upward. The analysis was also tuned so it eliminated the possibility that frogs were actually interested by the gross amount of worms or their movement and not their number.
The frog species involved in the experiment, Bombina orientalis or the Oriental fire-bellied toad (wrongly/named because it is a proper frog), a mildly toxic and colorful frog commonly used as a pet, was chosen because it belongs to a primitive taxonomic group (Archaeobactria) and thus offers the chance to study very old cognitive functions and proto-numerical abilities in lower vertebrates. The study involved seven adult frogs raised in captivity that were commonly fed with live worms every other day (only a moving worm triggers the predatory response). Before the experiment, the frogs were food-deprived for two days in order to increase their motivation.
After the two days of starvation, the frogs were exposed to their favorite prey in a free choice experiment: a frog was positioned between two groups of worms (the larval form of Tenebrio molitor, the mealworm beetle) where one group was greater in number than another. The tendency of the frog to move towards the greater number of worms was observed repeatedly.
When the frog had to choose between groups of one and two worms, it jumped towards the group with two worms 70% of the time. Between two and three, it chose three 60% of the time. Things got interesting after that: it seems that the frog cannot differentiate between three and four, four and five and so forth – its movements were ambiguous, with left and right jump rates of 50% and 50%. In all tests the p-values for estimates of jumping rates were below 0.05, and in some cases even below 0.01.
A confounding hypothesis is that the frog is merely integrating the total “worm area” that in its field of vision. What if the response was not stemming from “more individual worms” but from simply “more worm-like stuff”? To control for the greater visual area subtended by two worms versus one, the scientists covered parts of the worms, so the same area of worm surface area was visible in both groups. Even with this control, the frog consistently jumped 70% of the time towards the two worm group compared to the single worm group. So its brain is clearly able to perform a certain level of abstraction.
The frogs were also able to discriminate between large quantities such as 4 vs 8 worms, but in these cases the researchers were less sure whether this was due to true quantification or whether this was entirely an visual surface area effect.
The study was published on Aug 10, 2014 in Animal Cognition.