Cow traffic in an automatic milking rotary system : with or without parallel guiding of cows for milking

Abstract: Automatic milking systems and automatic milking rotary systems (AMR) where the cows are gathered and held in a waiting area prior to milking are dependent on cows to enter the milking unit voluntarily. A low rate of cows that voluntarily enter the milking unit affects both the economy, working safety and cow health negatively. Both the flow of animals as well as the interactions in this enclosed area can be affected by various designs of the waiting area. The present study examined dairy cow behaviour, milking time and labour in an AMR with two different waiting area designs: with parallel guiding or no parallel guiding. The parallel guiding consisted of two railings placed in front of the entrance to the AMR. These railings separated the cows, creating what could be seen as three corridors in front of the AMR. A control treatment without parallel guiding and a treatment with parallel guiding was applied for two weeks each. Two groups, G1 and G2, of approximately 60 lactating cows each of Swedish Red and Swedish Holstein breeds were included in the study. The G1 group consisted of mainly primiparous cows and G2 of mainly multiparous cows. For analysis of the behaviours, 20 focal animals were marked in each group. The staff gathered the cows in one of the two waiting area designs prior to milking but were not permitted to interact with the cows for 40 minutes after that. During these 40 minutes, the cow behaviour was studied. An ethogram was used. For example, it was noted if a cow pushed another cow. Both the performing and the receiving animal was noted. When 40 minutes had passed, the staff were allowed to fetch cows that had not entered the AMR. The period from the staff’s first interaction to the last cow entering the milking unit was also studied, each time staff interacted with a cow was noted as well at the total time for staff interaction. Data was collected during 6 milking occasions per group on the second week of each treatment. This was performed twice for each treatment, resulting in a total of 48 milking occasions studied, 24 during each treatment. The results showed a lower frequency of behaviours when parallel guiding was practised (p=0.04), as well as lower total staff interaction time (p=0.04) and staff interactions (p=0.0002). The majority of behaviours were performed by the lightest (40%) or heaviest animals (37%), and multiparous cows (58%). The number of performed, as well as received behaviours, seemed to be increasing with days in milk. The animal receiving a behaviour was in most cases primiparous cows (63%). The majority of the staff interactions were also performed on first parity cows (57%) and the heaviest cows (50%). The number of received staff interactions seemed to be increasing with days in milk. No difference was found between treatments in total time for milking, total walking and waiting time before milking or time spent in waiting area. Neither was a difference found regarding the time spent in waiting area as a group retrieved from the video recordings. However, the parallel guiding did overall not result in longer time intervals when compared to no parallel guiding. The milk yield during the periods when parallel guiding was practised was found to be higher compared to when no parallel guiding was practised (p<.0001). The parallel guiding was not found to have any negative effects on either working time, animal behaviour or milk yield. Further studies are needed to investigate the effects of the parallel guiding in more depth.