Effects of restraint and branding on rates and acoustic parameters of vocalization in beef cattle

An animal's vocalizations may be a useful reflection of its subjective stat e of welfare. We attempted to evaluate vocal response as an indication of a nimal distress. Calves (n = 189) were randomly allotted to four treatment g roups in a 2 X 2 factorial design, the factors being Branded vs. Not Brande d and Restrained vs. Not Restrained. On four consecutive days calves were b rought through a headgate and squeeze apparatus. Restrained calves were cau ght in the headgate for 3 to 5 s. The remainder were stopped at the headgat e but not restrained. On day 5, all calves were captured and restrained. Ha lf the animals were hot-iron branded and half were sham branded using an un heated iron. During branding, vocalizations were recorded. Digitized files were used to generate an audiospectrogram and a power spectrum for each cal l. We analysed 167 calls. During treatment, 65 calves vocalized. More brand ed than non-branded animals vocalized (58/95 compared with 7/94, P < 0.0001 ). Branded animals showed a greater frequency range in the fundamental, or lowest harmonic, of the audiospectrogram, (68.04 Hz +/- 5.33 compared with 28 Hz +/- 8.74, P < 0.05), a higher maximum frequency (186.66 Hz +/- 5.19 c ompared with 141.6 Hz +/- 6.6, P < 0.01). and a higher peak sound level (P < 0.05). The previous 4 days of restraint did not alter the probability of vocalizing, or any characteristics of the calls. This suggests that measuri ng vocal response may be particularly useful when the effects of relatively severe stressors are being investigated. The insensitivity of vocal respon se to moderate levels of stress may offer an advantage over cortisol or hea rt rate measurements which can reach high values in response to situational factors, leaving little capacity to respond to treatment. Analysis of voca lization may be among the more reliable and least invasive methods of asses sing acute distress in cattle. Based on findings of this study, we suggest that vocalization data should be interpreted as statistical properties of a treatment group, rather than indicating the condition of any individual an imal. (C) 1999 Elsevier Science B.V. All rights reserved.