Search for a command to run...
The objective of this study was to describe and compare the dynamics of milking behavior and milking performance during the first third of lactation of primiparous (PRI) and multiparous (MLT) Holstein (HO), Jersey (JE), and Holstein × Jersey crossbred (HJ) cows on a multibreed dairy farm with a batch milking system with automatic milking units. This retrospective observational study included information from 237,182 milking events in 790, 962, and 230 HO, JE, and HJ cows, respectively, from May 2023 to April 2024. The distribution of the 3 genetic groups in the herd was HO = 39.9%, JE = 48.5%, and HJ = 11.6%, which were maintained under the same management conditions in an organic-certified grass-fed herd in central Texas, United States. Data from 22 automatic milking units were used to monitor milking behavior and performance from calving until 90 DIM. The variables considered to assess milking behavior included milking unit kick-off (KO), incomplete milking (INC), teat cleaning failure (TCF), and teat not found (TNF). Continuous variables for the assessment of milking performance were milk yield (MY, kg/d), milking duration (MD, min), milking duration:milk yield ratio (MDY, min/kg), milk flow duration (MFD, min), mean milk flow rate (MMF, kg/min), and peak milk flow rate (PMF, kg/min). The first phase of the analysis compared the milking behavior and milking performance of PRI versus MLT cows within each of the genetic groups (HO, JE, and HJ), whereas the second phase compared the genetic groups within the parity categories (PRI and MLT). Binary variables (KO, INC, TCF, and TNF) were analyzed using logistic regression to assess potential associations of milking behaviors with parity category and genetic group (HO, JE, and HJ) in 3 lactation periods (<15 DIM, 15-60 DIM, and 61-90 DIM). Least squares means (SE) for each of the continuous variables (MY, MD, MDY, MFD, MMF, and PMF) were calculated by parity category and genetic group and compared using ANOVA for repeated measures analysis. Frequencies of KO, INC, TCF, and TNF, considered for assessing undesirable behaviors during milking, were consistently greater in PRI than MLT cows. In addition, MLT cows showed less variation in these variables, which, compared with those of PRI cows, remained more stable during the monitoring period. The comparisons among genetic groups identified the greatest MY in HO cows, whereas the results for the other performance traits were variable. Overall, the results from the current study highlight the differences in milking behavior during early lactation in PRI versus MLT cows among HO, JE, and their crosses. As anticipated, for most variables, the magnitude of these differences decreased as the lactation advanced, which may be associated with the process of habituation to milking during the first lactation.