Method for short-term prediction of milk yield at the quarter level to improve udder health monitoring

Abstract

Udder health problems are often associated with milk losses. These losses are different between quarters, as infected quarters are affected both by systemic and pathogen-specific local effects, whereas noninfected quarters are only subject to systemic effects. To gain insight in these losses and the milk yield dynamics during disease, it is essential to have a reliable reference for quarter-level milk yield in an unperturbed state, mimicking its potential yield. We developed a novel methodology to predict this quarter milk yield per milking session, using an historical data set of 504 lactations collected on a test farm by an automated milking system from DeLaval (Tumba, Sweden). Using a linear mixed model framework in which covariates associated with the linearized Wood model and the milking interval are included, we were able to describe quarter-level yield per milking session with a proportional error below 10%. Applying this model enables us to predict the milk yield of individual quarters 1 to 50 d ahead with a mean prediction error ranging between 8 and 20%, depending on the amount of historical data available to estimate the random effect covariates for the predicted lactation. The developed methodology was illustrated using 2 examples for which quarter-level milk losses are calculated during clinical mastitis. These showed that the quarter-level mixed model allows us to gain insight in quarter lactation dynamics and enables to calculate milk losses in different situations.

Milk losses linked to udder health treatments at dairy farms with automatic milking systems

“…Penry et al (2018) showed that there are significant associations between quarters for individual milking events, but they did not consider longitudinal data characteristics. Nevertheless, our earlier work (Adriaens et al, 2018) showed that both the peak yield and the persistency in unperturbed milk yield curves differ across front and hind quarters, and that the variability in quarter lactation shapes suggest that udder asymmetry and past infections may cause the expected milk yield of the left and right udder quarters to be different.…”

“…Quarter separation additionally provides a unique opportunity to study losses and dynamics in the different quarters separately and to distinguish local (i.e., related to the local inflammatory reaction and toxins) and systemic (i.e., related to the general sickness) effects of the infection (Adriaens et al, 2018).…”

Milk losses and dynamics during perturbations in dairy cows differ with parity and lactation stage

“…In this study we used the theoretical shape of the lactation curve (ULC) of AMS-milked dairy cows in the first 305 days to identify and characterize perturbations in the daily milk yield dynamics. To this end, we implemented an iterative, but computationally highly efficient method to estimate the presumed unperturbed lactation curve using the gamma function with three parameters (Wood model), as also proposed in Adriaens et al (2018, 2020) and Ben Abdelkrim et al (2019). Several other methods were tested for this study, including a 4 th order quartile regression model as proposed by Poppe et al (2020) and the Dijkstra lactation model (Dijkstra et al, 1997).…”

“…Wood’s model consists of 3 parameters, from which the a parameter mainly determines the scaling of the curve, whereas b and c specify the moment of peak production and slope. We preferred the Wood model over other lactation models because of its simplicity, its stability in the presence of missing data, the computational ease and its suitability for the purpose of this study (Adriaens et al, 2018; Ben Abdelkrim et al, 2019). To estimate the ULC that represents the milk production potential in absence of perturbations, an iterative fitting procedure was implemented to gradually remove milk yield data during perturbations.…”

“…This opens opportunities for precision phenotyping and targeted monitoring, starting from actual insight in lactation and disease dynamics, and including the aforementioned recovery and development rates and production losses. In order to study perturbations, first the theoretical unperturbed lactation curve has to be estimated, for example as proposed by Adriaens et al (2018), Ben Abdelkrim et al (2019), Adriaens et al (2020) and Poppe et al (2020). The deviations from this unperturbed curve can then serve as the basis for detecting and characterizing the deviations in milk yield and identification of challenges, especially in absence of detailed and reliable health registers.…”