Feeding behavior, milking behavior, and milk yields of cows milked in a parlor versus an automatic milking system

This study compared feeding and milking behavior and milk yields for cows housed in the same barn, fed the same ration, but milked with a conventional milking parlor (parlor) or automatic milking system (robot). Behavioral data were videotaped hourly 1 d/mo for 9 mo. Feeding behavior patterns differed and were more variable for parlor cows than for robot cows. Both groups had low feeding rates at night and early morning. Feeding activity increased after milking and feed delivery for parlor cows. Milking and feeding activity in the robot system increased after human intervention at 7 a.m.; feed bunk activity peaked 3 h later and remained relatively constant from 10 a.m. to 8 p.m. Percentages of cows at the feed bunk were significantly greater for robot cows than parlor cows only at 10 a.m. and 9 p.m. Batch milking of parlor cows with free access to feed, vs. sequential milking of robot cows, with restricted movement to feed by a one-way gate system, resulted in higher peak percentages of cows at the bunk for parlor cows. Lower, more consistent percentages of cows eating at one time suggests that less bunk space may be needed for cows in robotic milking systems. Higher percentages of cows were observed in the robot from 8 a.m. to 1 p.m. and again from 3 to 7 p.m. Percentages of cows in the robot holding area were greatest from 8 to 11 a.m. and 3 to 6 p.m. and were lowest from midnight to 6 a.m. Milk production over 39 d in summer for subsets of cows was slightly but significantly higher (26.4 vs. 25.8 +/- 0.2 kg/d) for cows in the robot group. Milking frequency, days in milk, parity, and maximum air temperature for 3 d (-2 d to day of observation) affected milk yield comparisons. Results have implications for design of feeding and handling facilities used with automated milking systems.