Bräucker R and Schwartzkopff J (1986): Frequency discrimination in the pigeon Columba livea.

In physiological studies, tones are mainly used as stimuli. Here frequency discrimination of pure sinusoidal sounds is compared with the frequency-resolving ability for sinusoidally frequency-modulated tones. The question is raised if the resolving power is improved by the latter stimulus procedure. The method used was heartrate conditioning. With the help of this classical conditioning technique, an associative connection between a conditioned (acoustic) stimulus ( = test tone) and a following unconditioned stimulus (a weak electric shock) is developed. In addition, a randomized number of reference tones which are not followed by a shock is played to the bird. The heart-beat rate recorded during the reference tones is compared with the heart-beat rate during the test tone, using the Student's t-test. The threshold criterion was determined at a onetailed error probability of 2.5%.
Four adult homing pigeons were tested in a sound-proof box. The tone duration was 3.5 s (20 ms rise-fall time). All signals had an intensity of 70 dB above the individual hearing threshold of the bird.
In a first experiment, the pigeons had to discriminate two pure tones at different frequency. Learning performance was better when the frequency of the reference tone was higher than the frequency of the test tone. The frequency of the reference tone was successively approximated to that of the test tone until difference learning could no longer be secured.
In the second experiment, the pigeons had to discriminate a pure tone as against a sinousoidally frequency-modulated tone of the same center frequency and varying depth of frequency modulation (modulating frequencies between 1 Hz and 25 Hz). The depth of frequency modulation was reduced until the change in the heart-beat rate was below criterion. This depth of modulation may be regarded as the frequency difference limen.