Broadcasting human voice to piglets (Sus Scrofa domestica) modifies their behavioural reaction to human presence in the home pen and in arena tests

The human-pig relationship develops through visual, tactile, olfactory, and auditory interactions. Our objective was to determine the effect of human voices on the development of the human-pig relationship. We hypothesised that human voice facilitates the development of the human-pig relationship. We studied the behaviour of 90 weaned female piglets divided into three treatments: human presence with (HPV) or without (HP) voice, and control (CTRL). For the HPV piglets, the experimenter was present (idle) for 5 min/day in the pen for three weeks and a female voice was broadcast from a speaker. The HP treatment was the same, but a recorded background noise was broadcast instead of the voice. For the CTRL piglets, only routine husbandry care was provided. Piglets were then tested twice in a 3x3 m test area alone for 5 min and then in the presence of the experimenter for 5 min. For test 1, the human voice was broadcast for HPV piglets and the background noise was broadcast for the others. In test 1, HPV and HP piglets investigated the experimenter earlier and more often than CTRL piglets (P < 0.05). HPV piglets moved sooner in the pen than CTRL piglets (P < 0.05). For test 2, only the background noise was broadcast; HPV piglets were thus deprived of the human voice they were used to. In test 2, HPV piglets expressed more stress reactions: their latency to move was longer compared to the others (P < 0.05). HPV piglets also had more physical and vocal interactions: they stayed in the experimenter area longer than HP and CTRL piglets (P < 0.05), and grunted more (P < 0.05). Finally we measured the time taken to move the pigs from their home pen to the truck to move to the farrowing building; we found no effect of the treatment (P < 0.05). In conclusion, broadcasting a human voice did not modify the pig response to human presence and handling in auditory conditions similar to the interaction sessions (i.e. test 1 and moving). However, not broadcasting human voice (test 2) induced stress responses and increased interactive behaviour, which suggests that piglets identified human voice as part of the experimenter’s necessary properties