The effect of premilking udder preparation on Holstein cows milked 3 times daily

Premilking udder preparation (including forestripping and duration of lag time-the time between first tactile stimulation and attachment of milking unit) might influence milking measures' such as milking unit on-time, incidence of bimodality; and milk flow rates in Holstein cows milked 3 times daily. Holstein cows (n = 786) from an 1,800-cow commercial dairy herd were enrolled under a restricted randomized design to determine the effect of 9 different premilking routines. Lag times were 0, 60, 90, 120, and 240 s and included forestripping or no forestripping for a total of 9 treatments (no forestripping for 0 lag time); the study was conducted from February to November 2008. All cow-treatment combinations were compared with the control: predipping plus forestripping and drying with 90 s of lag time. Cows were initially assigned to 1 of 3 treatments for a period of 7 d and upon completion of the first 7-d period were reassigned to a different treatment until all treatments had been completed. From one treatment period to the next, cows had to switch stimulation method with no restriction on lag time. Cows did not receive all treatments during the duration of the trial. Early- to mid-lactation cows (EML; 17-167 DIM) and late-lactation cows (LL; 174-428 DIM) were housed in 2 different pens. Milk yield was significantly different between dip + forestrip and dip + dry for 2 of the treatments for EML cows compared with dip + forestrip and 90 s of lag-time (DF90); however, this was not thought to be due to treatment because the significant lag times were very different (60 and 240 s) and neither was an extreme value. Milk yield did not differ with treatment for the LL cows. Milking unit on-time did not differ when comparing all treatments for EML with treatment DF90; however, an increase in milking unit on-time occurred when lag time was 60 s or less for LL cows. The highest incidence of bimodal milk curves was when lag time = 0 and this was independent of stage of lactation; a lag time of 240 s had the second-highest incidence of bimodal milk curves for EML and LL cows. Milk harvested in the first 2 min was lower for lag times of 0 and 240 s when compared with DF90. Increasing the lag time for all cows appeared to improve overall milking time efficiency (although lag time had no effect on EML cows).