Lactation curve modelling in dairy production: Applications at cow and herd level

Abstract

Various metrics have been proposed to evaluate milk production of dairy cows, like cumulative milk production over a specific period (e.g., 305 days, 365 days, or an entire lactation) and milk yield per day within a certain period. These metrics are based on simple calculations on raw data. However, they provide an overview of milk production performance without capturing changes in milk production over the entire lactation period. These changes, or patterns in milk production offer more information (e.g., peak yield, peak time, persistency) about the lactation, which can be useful for breeding and selection, health monitoring, and other applications. Lactation curve models can extrapolate and quantify lactation curves and estimate actual production from incomplete data sets, generating various lactation curve characteristics (LCC) to describe the curve in different ways. LCC can serve as a metric to evaluate milk production performance at the cow level and have diverse applications in various dairy research fields. However, some important research topics have received insufficient attention. This thesis was conducted to explore the application of lactation curve modelling based on farm data collected on commercial dairy farms in the Netherlands and Belgium. Applications in cow reproduction performance and herd economic performance were developed. Four objectives were formulated: (1) to predict lactation persistency for DIM 305 at different insemination moments, (2) to investigate the association between days post conception and persistency, (3) to summarize cow lactation curves into herd lactation curve characteristics (HLCC) and illustrate a field application of HLCC, and (4) to compare whether HLCC or the herd’s average 305-day milk production (HM305) is better able to explain herd economic performance.