Social hierarchy affects the adaption of pregnant sows to a call feeding learning paradigm

The aim of the study was to test whether adult sows are able to learn an individual acoustic signal for call-feeding in groups supplied with an electronic feeder. Further, we investigated whether and how the social rank of sows affects learning success. Thirty-six sows were examined in 6 successive trials. In each, the animals were kept together for establishing a social hierarchy a week before conditioning started. Agonistic interactions were observed and a dominance index (DI) was calculated for the sows of each trial. Based on the DI sows were categorised as (1) dominant, (2) subordinate, or (3) submissive. Afterwards groups were transferred to the experimental pen which was equipped with one electronic feeder supplemented with a loudspeaker and software, the call-feeding station (CFS). The training started with classical conditioning (7 days) where the animals entered the CFS spontaneously 6 times daily and received a portion of feed immediately after an individual acoustic signal had been played. In the following operant conditioning phase (13 days) the individuals had to learn that they could enter the CFS and receive feed only after they had heard their signal. The animals were called 6 times daily to feed the respective fraction of the daily feed allowance. On the average, after 8 days of operant conditioning the animals reached the learning criterion of following 80% of their calls. The success rates differed significantly between the three rank groups. In the dominant and subordinate groups 93% and 71% of the animals reached the learning criterion at the end of the experiment after 13 days of operant conditioning, while only 64% of the submissive sows did so. If only the number of successful, i.e. rewarded, enters of the station was considered those submissive animals who had reached the learning criterion did not differ significantly from the others. During learning, the time required to approach the CFS decreased significantly as well as the rate of false attempts to enter if another animal was called. At start of the operant training dominant sows blocked the entrance of the CFS. With increasing learning success of these sows this behaviour decreased significantly. The experiment has demonstrated that call feeding can be applied successfully with pregnant sows. It has the potential to increase animal welfare because, by calling them individually to the feeder, it provides the animals with a positive short time anticipation of unaffected feeding.