A method for measuring the aversiveness of sounds to domestic hens

Six hens' key-pecking at each of two simultaneously available perspex discs (keys) was maintained by intermittent access to wheat. Over a series of conditions one of four sounds (music, water-hose, poultry and train), at a range of decibel levels (90–100 dB(A)) was relayed to the experimental chamber when pecking was on one key, but not when it was on the other. The key associated with the sound was reversed twice during each 40-min session. The reversal was signalled by changing key-light colour.
The effect of the sounds on behaviour was assessed by measuring changes in the relative allocation of key pecks between the two keys (one with sound, the other without). A consistent effect of the sound was shown as a bias in both responding and time allocated to the keys. Bias could be seen in the behaviour measures for some sounds, and the biases owing to the presence of sound, were estimated mathematically from the generalized matching law. Generally, these noise biases increased with the increasing intensity of the sound and differed from sound type to sound type. Results showed that the sound of hens in a commerical poultry shed at 100 dB(A) was associated with the most bias and a piece of music at 90 and 95 dB(A) was associated with the least bias.
It is argued that the bias measures derived from this procedure provide quantitative estimates of the relative aversiveness of the sounds used. Such measures are important in evaluating environmental features that may affect hen welfare.