Transitioning dairy cows to automatic milking: Effects of different training methods on behavioral responses during...

Automatic milking systems (AMS) allow cows to choose when and how often they are visiting the milking unit, but introduction to these systems can be fear inducing. Sometimes cows are habituated to an AMS before milking is initiated, but there are no widely agreed upon training procedures for easing this transition. The aims of the current study were to: 1) assess behavioral responses associated with fear when cows were first introduced to AMS milking, and 2) test if these responses were affected by how cows were trained. We followed a single pen of 46 lactating cows (with an average parity of 2.8 ± 1.6) during a herd-wide transition from parlor to AMS milking. Cows were randomly assigned to either acclimation training regimen followed by baseline training (n = 26), or to just the baseline training (n = 21). Cows subjected to the acclimation regimen began training 13 d before the switch from parlor to AMS milking. This acclimation regimen consisted of a habituation phase that included exposure to a non-operational AMS unit, with cows encouraged (using positive reinforcement via a grain reward) to enter the milking unit. The acclimation training ended 3 d before the switch to AMS milking; at this point 11 of 21 of the acclimation-trained cows had freely entered the AMS and were thus included in our assessment of successfully trained cows. Baseline training was then applied to all cows in the pen once daily for the remaining 3 d; this involved farm staff moving cows individually into the non-operational AMS where they received a food reward. After the initiation of AMS milking, we monitored behavioral responses for their first 3 milkings for the cows receiving baseline training and the cows that successfully completed the acclimation training. The behavior responses assessed included time resisting when entering the milking unit (i.e., time stopped moving or moving backward at the entrance of the AMS), number of kicks during milking, and time spent eating during milking. We found that acclimation trained cows performed less resisting behavior across all events (model mean ± SE 0.6 ± 0.5 log s/milking) than did the baseline trained cows (1.8 ± 0.3 log s/milking), and resisting increased in the second and third milking relative to the first milking event (1st milking event: 1.0 ± 0.4 log s/milking, 2nd and 3rd milking events: 1.6 ± 0.4 log s/milking). Regardless of the training treatment, cows were similar in number of kicks and time spent eating in the AMS unit. We conclude that the acclimation training encouraged entrance into the AMS. We encourage future work to assess other training methods and potentially more sensitive response measures.