Abstract #T120
Section: Production, Management and the Environment (posters)
Session: Production, Management and the Environment 2
Format: Poster
Day/Time: Tuesday 7:30 AM–9:30 AM
Location: Exhibit Hall A
Session: Production, Management and the Environment 2
Format: Poster
Day/Time: Tuesday 7:30 AM–9:30 AM
Location: Exhibit Hall A
# T120
Effect of measures of milk yield and dry period length on prediction of milk loss in the subsequent lactation.
P. Pattamanont*1, M. I. Marcondes1,2, A. Bach3, J. S. Clay4, A. De Vries1, 1Department of Animal Sciences, University of Florida, Gainesville, FL, 2Department of Animal Science, Federal University of Vicosa, Vicosa, MG, Brazil, 3ICREA and Department of Ruminant Production, IRTA, Barcelona, Spain, 4Dairy Records Management Systems, North Carolina State University, Raleigh, NC.
Key Words: dry period, milk loss, prediction
Effect of measures of milk yield and dry period length on prediction of milk loss in the subsequent lactation.
P. Pattamanont*1, M. I. Marcondes1,2, A. Bach3, J. S. Clay4, A. De Vries1, 1Department of Animal Sciences, University of Florida, Gainesville, FL, 2Department of Animal Science, Federal University of Vicosa, Vicosa, MG, Brazil, 3ICREA and Department of Ruminant Production, IRTA, Barcelona, Spain, 4Dairy Records Management Systems, North Carolina State University, Raleigh, NC.
To optimize cows’ dry period lengths, it is important to estimate milk loss in the subsequent lactation associated with non-optimal days dry (DD). It is unclear how different measures of milk yield in the current lactation in combination with DD affect predicted milk loss. Therefore, the objective of this study was to quantify the effect of DD combined with 2 measures of milk yield in the current lactation on milk loss. DHI milk test records of 824,809 Holstein cows from 4,766 herds with the last dry date in 2014 or 2015 were obtained from DRMS in Raleigh, NC. Three groups of adjacent lactations were constructed: lactation 1 and 2 (lac1–2), lactation 2 and 3 (lac2–3), and lactation 3 and greater (lac3+). Included cows had ≥ 6 milk tests including fat and protein observations in the current lactation; gestation length of ≥ 270 d; DD from 5 to 120 d. Five models with a fourth-degree polynomial function of DD and control covariates with combinations of cumulative ECM yield to the last test day (ECMLTD), daily ECM yield 14 d before the dry off date (ECM14), and random herd-season effects were developed to predict 305-d ECM yield in the subsequent lactation (ECM305) for each group. Milk loss was calculated as the predicted ECM305 given a certain DD minus the predicted ECM305 at 58 DD at which ECM305 of all cows in these data was maximized. Compared with the model including only a polynomial DD function, the model including ECMLTD had a 10% smaller RMSE, whereas including both measures of milk yield reduced RMSE by 11%. The inclusion of the 3-way interaction reduced RMSE only by 0.2% compared with the model with only the 2-way interactions. The full model was selected to illustrate predicted milk loss of ECM305 by group. The effect of DD shorter than 58 d on milk loss was greatest in lac1–2. A low level of ECM14 resulted in less milk loss in all groups when DD was short, whereas a high level resulted in greater milk loss. In conclusion, milk yield interacts with DD and affects milk loss in the subsequent lactation. Both ECMLTD and ECM14 help predict milk loss and are therefore useful for economic optimization of DD.
Key Words: dry period, milk loss, prediction
